Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
STATUS OF THESIS
INTERRELATIONSHIPS BETWEEN USERS AND SYSTEM Title of thesis FLEXIBILITIES WITH PERCEIVED USABILITY OF ONLINE
AIRLINE RESERVATION SYSTEMS
ARIF MUSHTAQ
hereby allow my thesis to be placed at the Information Resource Center (IRC) of Universiti Teknologi PETRONAS (UTP) with the following conditions:
I. The thesis becomes the property of UTP
2. The IRC ofUTP may make copies of the thesis for academic purposes only.
3. This thesis is classified as
D Confidential
~ Non-confidential
If this thesis is confidential, please state the reason:
The contents of the thesis will remain confidential for _____ years.
Remarks on disclosure:
Endorsed by
Sign e of Author Signature o
Permanent address: Name of Supervisor
House 36A. Street 39B, Sector 1-9/4 Dr. Suziah Bt. Sulaiman
Islamabad, Pakistan
Date: ciO- /QJ. -dOt/ Date:
UNIVERSITI TEKNOLOGI PETRONAS
DISSERTATION TITLE: INTERRELATIONSHIPS BETWEEN USERS AND
SYSTEM FLEXIBILITIES WITH PERCEIVED USABILITY OF ONLINE
AIRLINE RESERVATION SYSTEMS
by
ARIF MUSHT AQ
The undersigned certify that they have read, and recommend to the Postgraduate Studies Programme for acceptance this thesis for the fulfilment of the requirements for the degree stated.
Signature:
Main Supervisor:
Signature:
Co-Supervisor:
Signature:
Head of Department:
Date:
Dr. Suziah Bt. Sulaiman
I
I "' ~ D. D. DOMINIC ~. "1',~ Professor
et:n•IPtll,., & luloiiildbi SO. as BapartmerC UnoYer .:G 1 ..,.,ologi PET~ONAS
Assoc. Prof Dr. P Dhana~-==dl6~~~non
Dr ~ .. ~ot1d Fec::..:il 8 Ha::-.s~-:rj 'l:>~j
Dr. Mohd Fadzil Bin Hassan
DISSERTATION TITLE: INTERRELATIONSHIPS BETWEEN USERS AND
SYSTEM FLEXIBILITIES WITH PERCEIVED USABILITY OF ONLINE
AIRLINE RESERV AT! ON SYSTEMS
by
ARIF MUSHT AQ
A Thesis
Submitted to the Postgraduate Studies Programme
in Fulfilment of the Requirement of the Degree of
DOCTOR OF PHILOSOPHY
IN INFORMATION TECHNOLOGY
UNIVERSITI TEKNOLOGI PETRONAS
BANDAR SERI ISKANDAR,
PERAK
DECEMBER 2011
DECLARATION OF THESIS
INTERRELATIONSHIPS BETWEEN USERS AND SYSTEM Title of thesis FLEXIBILITIES WITH PERCEIVED USABILITY OF ONLINE
AIRLINE RESERVATION SYSTEMS
I ARIF MUSHT A
hereby declare that the thesis is based on my original work except for quotations and
citations which have been duly acknowledged. I also declare that it has not been
previously or concurrently submitted for any other degree at UTP or other
institutions.
Witnessed by
Signature o
Permanent address: Name of Supervisor.
House 36A, Street 398, Sector I-9/4 Dr. Suziah Bt. Sulaiman
Islamabad, Pakistan
Date: do - I J - ol tl I I Date: ~-.").-+f/.:._1 .;}'--'{1-.do~i-1 ~~-
IV
DEDICATION
This dissertation is dedication to my father, who unconditionally loved me since I
was born I wish you could stay few more years to see all my success and amazing
accomplishments. I wish you could see that I am going to he a PhD. I wish you could
see that I have become a father ala loving daughter. You are truly missed. I love you'
v
ACKNOWLEDGEMENTS
It is a pleasure to thank those who made this thesis possible and supported me
throughout the course of my studies. Foremost, my eternal gratitude goes to my
loving wife, for her everlasting support and unconditional love. It would have been
next to impossible to write this thesis without her help and guidance. I am forever
grateful to my mom for her moral support and prayers. I would like to thank my
family for all their love and encouragement. I owe my deepest gratitude to my mother
in law and father in law who always remembered me in their prayers. I am thankful to
my beloved daughter "Manahil" who made me smile during the hectic research days.
To my supervisor, Dr. Suziah Bt. Sulaiman .. I am highly grateful for her help and
guidance in the research. I am forever obliged fi)r the encouragement and support that
you never hesitated to give me. Similarly, thanks to my co-supervisor Dr. P. D. D.
Dominic for his support and counseling. This work would not have been possible
without their guidance.
I owe endless thankfulness to all those such as students, staff, travel agents,
travelers and other anonymous evaluators who participated in the survey. Special
appreciation goes to Gehan for helping me in making the interface prototype. The
final study would have not been possible without them. Lastly, I am grateful to my
friend for providing me numerous distractions and relaxation opportunities during the
journey of this dissertation.
VI
ABSTRACT
It is very critical for the organizations to design flexible systems that are easy to
use and can accomplish all the requirements by way of offering customizability.
Philosophers argue that users are good in adapting the systems; however, research
shows users dissatisfaction with existing Online Airline Reservation Systems in terms
of task completion. Therefore, researchers are eager to find out ways for improving
online usability of the systems, how users' Perceived Usability of the system is
formulated by its flexibility functions. This research therefore examines travelers'
expectations, preferences and online behavior (Users' Flexibility) and aligns that with
designing of flexible online airline reservation systems (System's Flexibility) and
users' as evaluators of the online systems to determine its Perceived Usability through
users' effectiveness, efficiency and satisfaction (Perceived Usability).
In this dissertation, both quantitative and qualitative techniques were used to
analyze the data collected in the context of SF, lJF and PU of the systems. A redesign
solution for enhanced usability was developed based on HCI guidelines and the
flexibility tactics used in online travel agencies, which led to a proposed interface
with the integration of opaque mechanism. The two interfaces were used in the
experiment. Participants were requested to complete the evaluation of the existing and
proposed interfaces.
The findings suggested that users can be classified on the basis of their Flexible
Traveling Behavior which led to the development of a Users' Flexibility measuring
scale. It is further investigated that integration of opaque fares concept would increase
the usability of the system. Since flexibility is referred to its ability to respond to
internal or external changes, systems incorporated with opaque fares would serve the
role of external change agent by way of providing flexibility in users' decision
making and will also serve the role of internal change agent by way of providing the
capability of accepting changed decisions.
Vll
ABSTRAK
Ini adalah sangat penting bagi organisasi untuk mereka bentuk sistem yang
fleksibel dan mudah untuk digunakan serta boleh mencapai semua keperluan dengan
cara menawarkan kebolehan untuk mengubahsuai. Ahli-ahli falsafah berpendapat
bahawa pengguna berkebolehan untuk menyesuaikan diri menggunakan sesuatu
sistem, namun kajian menunjukkan rasa tidak puas hati pengguna dengan sistem yang
sedia ada dalam tempahan penerbangan secara talian dari segi menyelesaikan tugas.
Oleh sebab itu, para penyelidik amat berminat untuk mengetahui cara-cara untuk
meningkatkan kebolehgunaan system dalam talian, dan bagaimana persepsi pengguna
terhadap menganggap kebolehgunaan sistem dapat digubal menerusi fungsi
fleksibilitinya. Kajian ini meneliti jangkaan pelancong, keutamaan dan tingkah laku
mereka dalam talian (fleksibiliti pengguna) dan menjajarkannya dengan reka bentuk
system tempahan penerbangan dalam talian (sistem fleksibiliti) dan meletakkan
pengguna sebagai penilai sistem tersebut bagi menentukan kebolehgunaan melalui
kepuasan pengguna (persepsi kebolehgunaan).
Dalam disertasi ini, kedua-dua teknik kuantitatif dan kualitatif telah digunakan
untuk menganalisa data yang dikumpulkan dalam konteks fleksibiliti, kelonggaran
sistem pengguna dan kebolehgunaan sistem. Satu penyelesaian bagi mereka bentuk
semula untuk memberi kegunaan yang lebih tinggi telah dibangunkan berdasarkan
garis panduan HCI dan taktik yang fleksibel yang digunakan dalam agensi-agensi
pelancongan dalam talian. Satu rekabentuk sistem tempahan penerbangan dalam
talian yang baru telah diaplikasikan dan membawa kepada antara muka yang
dicadangkan dengan integrasi mekanisme legap. Kedua-dua antara muka telah
digunakan dalam eksperimen terse but. Para peserta telah diminta untuk melengkapkan
penilaian antara muka yang sedia ada dan yang dicadangkan.
Hasil penemuan mencadangkan supaya pengguna boleh dikelaskan berdasarkan
tingkah laku perjalanan mereka yang fleksibel yang lantaran itu membawa kepada
V111
pembangunan skala mengukur fleksibiliti seorang pengguna. Perkara ini disiasat
dengan lebih mendalam yang mana integrasi konsep tambang legap akan
meningkatkan kebolehgunaan system. Oleh scbab fleksibiliti dirujuk dengan
keupayaan untuk bertindak balas terhadap perubahan dalaman atau luaran, sistem
yang diperbadankan dengan tam bang legap akan ber peranan sebagai agen perubahan
luaran dengan menyediakan fleksibiliti supaya pengguna dapat membuat keputusan
dan juga akan berperanan sebagai agen perubahan dalaman melalui penyediaan
keupayaan menerima keputusan berubah.
!X
In compliance with the terms of the Copyright Act 1987 and the IP Policy of the university, the copyright of this thesis has been reassigned by the author to the legal entity of the university,
Institute of Technology PETRONAS Sdn Bhd.
Due acknowledgement shall always be made of the use of any material contained in, or derived from, this thesis.
© ArifMushtaq, 2011 Institute of Technology PETRONAS Sdn Bhd All rights reserved.
X
TABLE OF CONTENTS
STATUS OF THESIS ............................................................................................. .
APROVAL PAGE ................................................................................................... . II
TITLE PAGE ............................................................................................................ 111
DECLARATION....................................................................................................... 1v
DEDICATION..................................................................... .................................... v
ACKNOWLEDGEMENTS.................................................................. ................... v1
ABSTRACT ............................................................................................................. VII
ABSTRAK ................................................................................................................ viii
COPYRIGHT PAGE............................................................................................... X
TABLE OF CONTENTS......................................................................................... x1
LIST OFT ABLES ................................................................................................... xvii
LIST OF FIGURES ................................................................................................... xix
LIST OF ABBREVIATIONS .................................................................................. xxi
Chapter
I. INTRODUCTION........................................................................................ I 1.1 Research Background.......................................................................... I 1.2 Problem Statement............................................................................... 4 1.3 Research Aims and Objectives............................................................ 6 1.4 Research Methodology........................................................................ 7 1.5 Scope of Research ......... ............. ... .............. ........................ ................ 9 1.6 Research Contribution......................................................................... I 0 1.7 Organization of the Thesis................................................................... 12
2. LITERATURE REVIEW..... ... ... ...... ..... .......... .. ........ ...... ... .... ....... ... ....... 13 2.0 Chapter Overview................................................................................ 13 2.1 Flexibility and Online Airline Reservation Systems........................... 13
2.1.1 Flexibility Concepts................................................................ 13 2.1.1.1 System's Flexibility ............. ..... ... ................... ....... ..... 14 2.1.1.2 Users' Flexibility......................................................... 16 2.1.1.3 Conceptual Linking between SF and UF ....... ....... ...... 18
2.1.2 Airline Reservation Systems................................................... 20 2.1.2.1 Computer Reservation Systems ..... ............................. 20 2.1.2.2 Self-Booking Tools vs. Online Travel Agencies........ 23
2.2 Usability of Online Airline Reservation Systems................................ 31 2.2.1 Usability Perception......................... ...................................... 32
2.2.1.1 Usability Perception by Performing Content Analysis 33
Xl
2.2.1.2 Usability Perception through User's Internet Adoption33 2.2.1.3 Usability Perception based on Users' Preferences and Expectations ......... .. ....... ................... .. .. .. ......... .. ....... ....... ...... 34 2.2.1.4 Usability Perception based on Online Behaviour of Web Users............................................................................. 34
2.2.2 Conceptual Linking between PU and FTB and FOARS....... 35 2.3 Usability vs. Flexibility...................................................................... 36
2.3.1 Opaque Selling Mechanics in System's Flexibility.............. 37 2.3.2 Opaque Selling Mechanics in User's Flexibility................... 39
2.4 Chapter Summary .............................................................................. 40
3. METHODOLOGY...................................................................................... 43 3.0 Chapter Overview.............................................................................. 43 3 .I Research Methodology . . . . . . .. .. . . . . . . . . . . . . . . . . . . .. .. . . . . . . . .. .. . . . . . . . . . . . . . . . . . . .. .. . . . . 4 3
3.1.1 Phase I: Assessing Users' Needs........................................... 48 3.1.2 Phase II: Classification of Users .......................................... 48 3.1.3 Phase III: Case Study............................................................ 49
3.2 Phase 1: Systems' Flexibility & Users' Flexibility............................ 49 3 .2.1 Study I: Issues with Flexibility............................................ 49
3.2.1.1 Rationale.................................................................... 50 3.2.1.2 Methodology............................................................. 51 3.2.1.3 Validity...................................................................... 52 3.2.1.4 Sample Size............................................................... 52 3.2.1.5 Response Rate........................................................... 53 3 .2.1.6 Scale .......................................................................... 53 3.2.1.7 Analysis..................................................................... 53
3.2.2 Study 2: Users' Flexible Behavior in Terms of Compromising on SQA.................................................................................. 54 3 .2.2.1 Rationale.................................................................... 54 3.2.2.2 Methodology............................................................. 54 3.2.2.3 Validity...................................................................... 55 3 .2.2.4 Sample Size............................................................... 56 3.2.2.5 Response Rate........................................................... 57 3.2.2.6 Scale.......................................................................... 57 3.2.2.7 Analysis .................................................................... 57
3.2.3 Study 3: Integration of OTA Features can make SBTs more FOARSs ................................................................................ 58 3.2.3.1 Rationale..................................................................... 58 3.2.3.2 Methodology............................................................ 58 3.2.3.3 Analysis..................................................................... 59
3.3 Phase II: Users' Classification and Interrelationship Testing of Variable.............................................................................................. 60 3.3.1 Study 4: Users' Perception on Factors Influencing FTB ...... 60
3.3.1.1 Rationale.................................................................... 60 3 .3 .1.2 Methodology . .. .. .. . . . . . . . . . . . . . . .. . . .. .. . . . . . . . . . . . .. .. . . . . . . . . . .. . . . .. . 60 3.3.1.3 Validity...................................................................... 62 3.3.1.4 Sample Size ............................................................... 62 3.3.1.5 Response Rate........................................................... 64
Xll
3.3.1.6 Scale ........................................................................... . 3.3.1.7 Analysis ............................................................... .
3.3.2 Study 5: Study to Classify Users' on the Basis ofFTB ......... . 3.3.2.1 Rationale .................................................................... . 3.3 .2.2 Methodology .............................................................. . 3.3.2.3 Validity ....................................................................... . 3.3 .2.4 Sample Size ................................................................ . 3.3.2.5 Response Rate ............................................................ . 3.3.2.6 Scale ........................................................................... . :U.2.7 Analysis ...................................................................... .
3.3.3 Research Way Forward .......................................................... . 3.4 Phase III: Case Study to Test the Proposed Framework ..................... .
3 .4.1 Rationale ................................................................................. . 3.4.2 Methodology .......................................................................... . 3.4.3 Validity ................................................................................... . 3.4.4 Sample Size ............................................................................ . 3.4.5 Response Rate ........................................................................ . 3.4.6 Paper Prototype ...................................................................... . 3.4.7 Usability Evaluation of Prototype .......................................... . 3.4.8 Scale ....................................................................................... . 3.4.9 Analysis .................................................................................. .
3.5 Statistical Methods .............................................................................. . 3.5.1 Descriptive Analysis .............................................................. . 3.5.2 Pearson Coefficient Correlation ............................................. . 3.5.3 Reliability Analysis ................................................................ . 3.5.4 ANOV A ................................................................................ . 3.5.5 ANCOVA ............................................................................... . 3.5.6 MANOVA .............................................................................. . 3.5.7 Multiple Regressions .............................................................. . 3.5.8 Post Hoc Scheffe's Test ......................................................... . 3.5.9 Kendall's Tau ......................................................................... .
3.6 Chapter Summary ............................................................................... .
4. RESULTS AND ANALYSIS ................................................................ . 4.0 Chapter Overview ............................................................................... . 4.1 Phase I: User Needs Associated with System's Flexibility ................ .
4.1.1 Descriptive and Reliability Analysis ...................................... . 4.1.2 Hypotheses Testing H 1 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••.
4.1.3 Hypotheses Testing H2 ..•..............•......•.•..•...•..•......•..•.............
4.1.4 Hypotheses Testing H3 .•..........................................................
4.1.5 Hypotheses Testing H4 .................•......•.•..•...•..•.......................
4.1.6 Hypotheses Testing H5 ........................................................... .
4.2 Phase 1: Users' Flexibility in Terms of Compromising on SQAs? ..... . 4.2.1 Assumptions in ANOVA to Test H6 ..•...................................
4.2.1.1 Box-and-Whisker Plot H0 .•.....•.•......•..•..•....................
4.2.1.2 Means Plot H6 ............................................................ .
4.2.1.3 T-TestH6 ..........................•......•......•.•..•..•..•...•.............
4.2 .1.4 Error Bars H6 .............................................................. .
Xlll
64 64 65 65 66 66 66 67 67 67 70 72 72 73 74 74 75 75 76 76 76 78 78 79 79 79 80 80 81 81 81 82
83 83 83 83 85 86 87 88 88 89 89 89 90 91 91
4.2.2 Hypothesis Testing H6 ........................................................... 93 4.2.2.1 One-way Analysis of Variance H6 ............................ 93 4.2.2.2 Post-hoc Scheffe Tests H6 ......................................... 94 4.2.2.3 Effect Size for On~-way ANOV A H6 ....................... 94
4.2.3 Assumptions in ANOV A to Test H7 ...•................................. 95 4.2.3.1 Box-and-Whisker Plot H7.......................................... 95 4.2.3.2 Means Plot H7....•....•.•....•.•.•.•.•.•...••..•.•.•.••.•..•.•..•........ 96 4.2.3.3 T-Test H7 •..•.•..•....••.•..•.•.•...•.•.•.........•.•.•.••.•..•.•....•..•.•. 97 4.2.3.4 Error Bars H7 ..•.••....•.••.•.•.•.•.•••..•.•.•...•.••.•.••....•.•..•....•. 97
4.2.4 Hypothesis Testing H7..........................•................................ 98 4.2.4.1 One-way Analysis of Variance H7 .•.•..•.•.••.•.••.•..•.•..•. 99 4.2.4.2 Post-hoc Scheffe Tests H7 .••••.•.•.•.•.•.•..•.•.••.•.•..•.••.•..•. 99 4.2.4.3 Effect Size for One-way ANOV A H7 •.••.•....•..•.•....... l 00
4.3 Phase 1: Users' Satisfaction with SBTs against their rated OTA Feature . . . . . . . . . . . . . . .. . . .. .. . . . . . .. . . . .. . . . .. . . .. . . . .. . . . . . . . . . . . . . . . . .. . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . l 00 4.3.1 Hypotheses Testing Hs ..... ..................................................... I 01
4.3.1.1 Means Plot Hs ........................................................... 101 4.3.1.2 Two-way Analysis ofVariance H8 ............................ 102
4.4 Phase II: Users' Perception on Flexible Traveling Behavior ............ 104 4.4.1 Qualitative Analysis .............................................................. 104
4.5 Phase II: Classification of Users on the Basis of Their FTB ............. 109 4.5.1 Hypotheses Testing H9 .......................................................... 109 4.5.2 Hypotheses Testing H10 ......................................................... Ill 4.5.3 Hypotheses Testing H11 ......................................................... Ill 4.5.4 Hypotheses Testing H12 ......................................................... 114
4.5.4.1 One-way Analysis of Variance H12 ........................... 114 4.5.4.2 Analysis of Covariance H12 ....................................... 115
4.6 Phase III: Users' Flexibility is determined by SQAs and EVs .......... 116 4.6.1 Hypotheses Testing H13 ......................................................... 116
4.6.1.1 Scatter Plots H13 ........................................................ 116 4.6.1.2 Pearson Correlation Coefficients H13 ........................ 117 4.6.1.3 Multiple Regression Analysis Hu ............................. 118 4.6.1.4 Analysis ofVarian~e H13 ........................................... 118 4.6.1.5 Testing and Interpreting Model Coefficients H13 ...... 119
4.7 Phase III: Perceived Usability with Existing and Proposed Systems 120 4.7.1 Hypotheses Testing H14 ......................................................... 120
4.7.1.1 Descriptive Statistics on Effectiveness H14 ............... 120 4.7.1.2 Two-way Analysis of Variance on Effectiveness H14 121 4.7.1.3 Error Bars on Effectiveness H14 ................................ 121 4.7.1.4 Interaction Effect on Effectiveness H14 ..................... 122 4.7.1.5 Descriptive Statistics on Efficiency H14 .................... 123 4.7.1.6 Two-way Analysis of Variance on Efficiency H14 .... 124 4.7.1.7 Error Bars on Efficiency H14 ..................................... 124 4. 7 .1.8 Interaction Effect on Efficiency H14.......................... 125 4. 7 .I. 9 Descriptive Statistics on Satisfaction H14.................. 126 4.7.1.10 Two-way Analysis of Variance on Satisfaction H14 127 4.7.1.1 I Error Bars on Satisfaction H14 ................................. 127 4.7.1.12 Interaction Effect on Satisfaction H 14 ..................... 129
XIV
4.7.1.13 Levene's Test on Existing Systems 1-! 14 .................... 129 4.7.1.14 Post-hoc Comparisons on Existing Systems 1-! 14 ....... 130 4.7.1.15 Levene's Test on Proposed Systems H 14 ................... 130 4.7.1.16 Post-hoc Comparisons on Proposed Systems H 14 ..... 130
4.8 Phase lll: Effect of Users' Flexibility on Proposed Systems? ............. 131 4.8.1 Hypotheses Testing H1s ........................................................ 131
4.8.1.1 Scatter Plots H1s ......................................................... 131 4.8.1.2 Homogeneity of Covariance's H1s .............................. 133 4.8.1.3 Alpha Adjustment l-!1s.......... .. ................................... 135 4.8.1.4 Univariate ANOVAs 1-! 15 ........................................... 135
4.8.2 Hypotheses Testing H1 6 ........................................................... 137 4.8.2.1 Descriptive Statistics on Perceived Usability H16 ....... 137 4.8.2.2 Levene's Test of Equality of Error Variances H16 ...... 138 4.8.2.3 Alpha Corrections for Post-hoc Comparisons H16 ...... 139 4.8.2.4 Post-hoc Multiple Comparisons 1-1 16 ........................... 139
4.9 Chapter Summary................................................................................ 143
5. DISCUSSION............................................................................................... 144 5. 0 Chapter Overview................................................................................ 144 5.1 System'sFlexibility ............................................................................. 147 5.2 Users' Flexibility in terms of Compromising on SQAs ofOARSs..... 148 5.3 Integration Assessment of OTAs Features into SBTs ......................... 150 5.4 Users' Perception on Factors Influencing Upon FTB ........................ 152 5.5 Classification of Users on the Basis of Their FTB ............................. 159 5.6 Role of SQAs and External Variables in Flexible Behavior of Traveler 160 5. 7 Users' Perceived Usability with Existing and Proposed Systems ...... 162 5.8 Multivariate Main Effect of Users' Flexible Traveling Behavior on
Proposed System's Effectiveness. Efficiency and Satisfaction........... 164 5.9 Recommendations for FOARS ........................................................... 164
5.9.1 Graphical User Interface for FOARS ...................................... 167 5.9.1.1 Home Page ofFOARS ................................................ 167 5.9.1.2 Booking Interface ofFOARS...................................... 168 5.9.1.3 Flight Search Window for Flexible Travelers............. 168 5.9.1.4 Travelers' Details Window for Flexible Travelers...... 169 5.9.1.5 Payment Window for Flexible Travelers.................... 169 5.9.1.6 Immature Ticket Window for Flexible Travelers ..... 170 5.9.1. 7 Flight Search Window for Int1exible Travelers.......... 170 5.9.1.8 Travelers' Details Window for Inflexible Travelers... 171 5.9.1.9 Payment Window for Inflexible Travelers.................. 171 5.9.1.1 0 Ticket Window for Inflexible Travelers.................... 172
5.10 ChapterSummary ............................................................................... 172
6. CONCLUSIONS........................................................................................... 173 6.1 Dissertation Summary.......................................................................... 173 6.2 Research Findings and Contributions.................................................. 175
6.2.1 Objective I .............................................................................. 175 6.2.2 Objective 2 ... .. . .... ...... .......... ... ..... .. . ..... .............. .................. 176 6.2.3 Objective 3 ....... ... . ...... ... ..... ... ....... ....... ........... ... ............... 177
XV
6.3 Future Work ....................................................................................... 178
REFERENCES .................................................................................................. 179
LIST OF PUBLICATIONS............................................................................... 195
APPENDICES
A. Survey on Evaluation of Online Airline Reservation Systems .................... 196
B. UFB in terms of Compromising on Service Quality Attributes ................... 199
C. User's Flexibility towards Flexible Online Airlines Reservation Systems .. 202
D. Transformation of Evaluators Users' Flexibility Scoring ............................ 204
E. Classification of Travelers' on the Basis of their FTB ................................. 205
F. Permission to Collect Data from Travel Agents for Research Purposes ...... 206
G. Permission to Collect Data from Travel Agents at MATTA fair 2011 ........ 207
H. Usability Evaluation Checklist ..................................................................... 208
XVI
LIST OF TABLES
Table 2.1 : Definitions of Flexibility in the Context of System Engineering ......... 15
Table 2.2 :Comparison Chart on Offered Features by SBTs and OTAs.. .............. 26
Table 3.1 :Research Questions and Hypothesis ..................................................... 45
Table 3.2 :Emerging Themes on Factors Influencing upon FIB and PU .............. 65
Table 3.3 :Uni-directional Scale to Measure Users' Flexibility ............................. 68
Table 3.4 :Transformation Scoring Scale ............................................................... 69
Table 4.1 :Coefficient of the Factors Affecting Usability of Online Systems ....... 84
Table 4.2 :Flexibility of Existing Systems ............................................................ 87
Table 4.3 : T-Test on User's Flexibility with SBTs ................................................ 91
Table 4.4 : One-Way AN OVA on Satisfaction Level with Existing SBTs ............ 94
Table 4.5 : Multiple Comparisons on Satisfaction Level with Existing SBTs ........ 94
Table 4.6 : T-Test on User's Flexibility with Existing OTAs ................................. 97
Table 4. 7 : ANOV A on Satisfaction Level with Existing OT As ............................ 99
Table 4.8 : Multiple Comparisons on Satisfaction Level with Existing OTAs ....... I 00
Table 4.9 :Mean & SD on the Recommendation of Integrating OTA Features ..... 102
Table 4.10 : Two-way AN OVA on the Satisfaction Level with Existing OARSs. I 03
Table 4.11 : Demographics of Online Travel Forums Respondents ......................... I 04
Table 4.12: Results of Online Travel Forums .......................................................... I05
Table 4.13: Demographics of in-depth Interview .................................................... 106
Table 4.14: Results of in-depth Interviews .............................................................. 107
Table 4.15 : Demographics of Focus Group Interviews ........................................... I 07
Table 4.16: Results of Focus Group ......................................................................... l08
Table 4.17 : Users' Flexibility Transforming Scoring Scale ..................................... I 09
Table 4.18: Range for Classification ofUFB on the basis ofUFS .......................... 110
Table 4.19 : Interrelationship b/w UFB & their Perceived Usability of FOARS ..... 111
Table 4.20: Pearson Correlations on Interrelationship between UFB and SF .......... l13
Table 4.21 : One Way Independent AN OVA with PU and System's Flexibility ..... 114
Table 4.22: ANOVA with PU and System's Flexibility .......................................... 115
Table 4.23: ANCOVA by Introducing System's Flexibility as Covariate ............... l15
Table 4.24 : Correlations b/w Flexible Personality, SQA and External Variables ... 118
XVII
Table 4.25: Multiple Correlation Analysis b/w EV, SQA and Users' Flexibility .... 118
Table 4.26: AN OVA with External Variables and SQAs ........................................ 119
Table 4.27: Values of Coefficients and T-test.. ........................................................ 119
Table 4.28: Descriptive Statistics of Effectiveness .................................................. 120
Table 4.29: Two-way AN OVA of Effectiveness ..................................................... 121
Table 4.30: Descriptive Statistics of Efficiency ....................................................... 123
Table 4.31 :Two-way ANOVA of Efficiency .......................................................... 124
Table 4.32 : Descriptive Statistics of Satisfaction ..................................................... 127
Table 4.33 : Two-way ANOV A of Satisfaction ........................................................ 127
Table 4.34: Levene's Test on Existing Systems ....................................................... 129
Table 4.35: Multiple Comparison to Test Users' Classification for ES ................... 130
Table 4.36 : Levene's Test on Proposed Systems ..................................................... 130
Table 4.37: Multiple Comparison to Test Users' Classification for PS ................... 131
Table 4.38: Box's Test of Equality of Covariance Matrices ..................................... l33
Table 4.39: Multivariate Tests .................................................................................. l34
Table 4.40 : Univariate ANOV As ............................................................................. 136
Table 4.41 :Descriptive Statistics on PU .................................................................. 137
Table 4.42: Levene's Test of Equality of Error Variances ....................................... 139
Table 4.43: Post-hoc Multiple Comparisons using Scheffe ..................................... 141
Table 5.1 :Summary of Research Questions, Hypotheses and Results .................. 145
XVlll
LIST OF FIGURES
Figure 2.1 : Two Different Perspectives of Flexibility ........................................... 14
Figure 2.2 : Different Elements of Human Computer Interaction ........................... 16
Figure 2.3 : Flexibility as the Multiplicity of Ways for Information Exchange ...... 16
Figure 2.4 : Building Construct of Cognitive Psychology ...................................... 17
Figure 2.5 : Conceptual Linking between SF and UF ........................................... 19
Figure 2.6 : Distribution Channels for Airline Reservation Systems ....................... 22
Figure 2.7 :U.S. Travel Market by Channel, 2008 and 2011 .................................. 25
Figure 2.8 : Lowest Airfare Search Results through Matrix Display ....................... 27
Figure 2.9 :Alternative Airport Search to Find the Lowest Possible Fare .............. 27
Figure 2.10: Hotel Search, Results Display & Sorting .............................................. 28
Figure 2.11: Opaque Fares Offered by Priceline for Discounted Tickets ................. 29
Figure 2.12: Opaque Fares Offered by Hotwire for Discounted Tickets .................. 30
Figure 2.13: Flexible Date Search Results by Hot wire ............................................ 30
Figure 2.14: Ticket Reservation through Intermediate ............................................. 38
Figure 3 .I : Research Methodology ......................................................................... 44
Figure 3.2 : Research Model for the Hypothesis to be Tested ................................. 51
Figure 3.3 :User's Analysis of the Participants from UTP ...................................... 53
Figure 3.4 :User's Analysis of Participants from Travel Agencies ......................... 56
Figure 3.5 : User's Analysis of Participants from UTP for Study 2 ........................ 56
Figure 3.6 :User's Analysis of Participants from Online Travel Forums ............... 63
Figure 3.7 :User's Analysis of Participants from In-depth Interviews .................... 63
Figure 3.8 :User's Analysis of Participants from Focus Group .............................. 63
Figure 3.9 :User's Analysis of Participants for Transformation Scale .................... 67
Figure 3.10: Proposed FOARS after Integrating Opaque Fair into SBTs ................. 71
Figure 3.11: Increased Usability by Improving SF and Users Decision ................... 71
Figure 3.12: Prototype of a Flexible Booking Window for Flexible Travelers ........ 75
Figure 3.13: Prototype of a Flexible Booking Window for Inflexible Travelers ...... 75
Figure 4.1 : Box Plot showing Satisfaction Level with Existing SBTs ................... 89
Figure 4.2 :Means Plot on Satisfaction Level with Existing SBTs ......................... 91
Figure 4.3 :Error Bar on Satisfaction Level with Existing SBTs ............................ 92
XIX
Figure 4.4 : Box Plot on Satisfaction Level with Existing OTAs.. .......................... 95
Figure 4.5 : Means Plot on Satisfaction Level with Existing OTAs.. ...................... 96
Figure 4.6 :Error Bar on Satisfaction Level with Existing OTAs ........................... 98
Figure 4.7 :Means Plot on the Recommendation oflntegrating OTA Features .... 101
Figure 4.8 :Data Distribution after Classifications of Users' .................................. 110
Figure 4.9 :Scatter Plot between Users' Flexibility and SQAs ............................... 116
Figure 4.10: Scatter Plot between Users' Flexibility and External Variables ........... 117
Figure 4.11: Effect of User Classification of FB on Effectiveness ........................... 122
Figure 4.12: Effect of Existing & Proposed Systems on Effectiveness .................... 122
Figure 4.13: Interaction Effect of Existing & PS on Effectiveness .......................... 123
Figure 4.14: Effect of User Classification ofFB on Efficiency ................................ 125
Figure 4.15: Effect of Existing and Proposed Systems on Efficiency ...................... 125
Figure 4.16: Interaction Effect of Existing & Proposed Systems on Efficiency ...... 126
Figure 4.17: Effect of User Classification of Flexible Behavior on Satisfaction ...... 128
Figure 4.18: Effect of Existing and Proposed Systems on Satisfaction .................... 128
Figure 4.19: Interaction Effect of Existing and PS on Satisfaction .......................... 129
Figure 4.20: Linear Relationship between Eff<!ctiveness and Efficiency ................. 132
Figure 4.21: Linear Relationship between Effectiveness and Satisfaction ............... 132
Figure 4.22: Linear Relationship between Efficiency and Satisfaction .................... 132
Figure 4.23: Bar Chart Showing Effectiveness of the Proposed System .................. 137
Figure 4.24: Bar Chart Showing Efficiency of the Proposed System ....................... 138
Figure 4.25: Bar Chart Showing Satisfaction of the Proposed System ..................... 138
Figure 5.1 :Comparison of Existing and Proposed Reservation Approaches ......... 166
Figure 5.2 :Home Page for Flexible Online Airline Reservation System ............... 167
Figure 5.3 :Booking Window for Flexible Online Airline Reservation System .... 168
Figure 5.4 :Flight Search Window for Flexible Travelers ...................................... 168
Figure 5.5 :Travelers' Details Window for Flexible Travelers ............................... 169
Figure 5.6 :Payment Method for Flexible Travelers ............................................... 169
Figure 5.7 :Immature Ticket for Flexible Travelers ................................................ 170
Figure 5.8 : Flight Search Window for Inflexible Travelers .................................... 170
Figure 5.9 :Travelers' Details Window for Inflexible Travelers ............................ 171
Figure 5.10: Payment Method for Inflexible Travelers ............................................ 171
Figure 5.11: Confirmed Ticket for Inflexible Travelers ........................................... 172
XX
ARS
ACM
B2C
CRS
CI
FIB
FOARS
FG
FR
GDS
GUI
HCI
ID
NFR
OTA
OCBT
OARS
OTF
PU
PLF
SBT
SF
SABRE
SIGCHI
SQA
TAM
TPB
UF
UFS
LIST OF ABBREVIATIONS
Airline Reservation System
Association for Computing Machinery
Business to Consumer
Computer Reservation System
Confidence Interval
Flexible Traveling Behavior
Flexible Online Airline Reservation System
Focus Group
Functional Requirements
Global Distribution System
Graphical User Interface
Human Computer Interaction
In-depth
Non-Functional Requirements
Online Travel Agencies
Online Corporate Booking Tool
Online Airline Reservation System
Online Travel Forums
Perceived Usability
Passenger Load Factor
Self-booking Tool
System Flexibility
Semi-Automated Business Research Environment
Special Interest Group on Human Computer Interaction
Service Quality Attributes
Technology Acceptance Model
Theory of Planned Behavior
User Flexibility
User Flexibility Score
XXI
1.1 Research Background
CHAPTER I
INTRODUCTION
Airline Reservation Systems are the computerized systems that are used for storing
and retrieving information in order to conduct air travel related transactions [I]. Most
of the airlines have their own Self-booking Tools (SBTs) also referred to as Online
Corporate Booking Tools (OCBTs) that provide an opportunity to their clients to
make online reservations [2]. Any sale made through airline's offices directly or
through their SBTs is referred to as direct sale. On the other hand, Global Distribution
Systems (GDSs) connects airline offices with the offline and Online Travel Agencies
(OTAs) [3]-[5]. GDSs book and sell tickets for multiple airlines. Any sale made
through GDSs is referred to as sale through intermediaries.
Airlines opt for selling tickets through direct channels and also through
intermediaries. However, in order to differentiate their reservation channel from
others and to increase the direct sale, airlines have invested heavily in deploying a
range of tactics, such as, featuring their web site URLs across the marketing and
advertising communications, Web fares, reward mileage bonuses, and negative
incentives for non-preferred booking channels [6]. In addition to this many airlines
also dispose off their distressed inventory by providing last-minute sale discounts in
order to secure incremental revenue, where airline offers its unsold inventory at
heavily discounted prices before it perishes. This selling approach is adopted because
it does not disrupt the existing distribution channels or retail pricing structure [7] and
becomes a productive source of incremental revenue for the airline. The best part of
this form of selling is that, although travelers enjoy highly discounted fares, they do
not have to make predictions or face extremities in predicting specifies of their
traveling itineraries. For example, Malaysian airline has launched last minute flights
m 2008 to increase the revenue and average load factor [8]. According to the
managing director of Malaysia Airline, Datuk Seri Idris Jala "the everyday low fares
will create new demand for people who do nor fly with Malaysia Airline", hence
potentially increasing the average passenger load factor from 70% on each flight by
filling up unsold seats. Besides all these tactics made by the airlines, however, more
than 50% sale is done through GDSs [9]. Furthermore, researchers [I 0] reported that
this type of direct selling at the last-minute could be very risky for the airline, since
the potential travelers may prefer waiting for last-minute sales and not purchase in
anticipation of heavy discounts [10]. Such a condition may put an airline in a very
risky position with potential possibility of rev,~nue loss. That is why this practice is
substantially criticized by analysts and researchers, who refer to it as a vivacious cycle
of price degradation that can eventually destroy the airlines [II].
The second way to dispose-off distressed inventory is through opaque channels.
The term 'opaque inventory' indicates selling of unsold travel inventory at heavily
discounted price and it is called as being opaque, because some of the attributes ofthe
service supplier such as, name in the case of airline or hotel etc., are kept hidden and
only revealed to the traveler once the purchase has been materialized [12]. It is called
opaque selling because of its innovative mechanisms for marketing and price
discrimination [13], [14]. Opaque inventory selling is like a box, full of surprises and
travelers who are interested in buying opaque inventory products are high in price
sensitivity and low in specifies of travel plan. Thus this form of selling immediately
captures the attention of travelers who would like to keep their travel expenses within
limited budgets. As for the airlines, in order to minimize effect of price degradation
on their revenue, accepted the role played by opaque selling intermediaries, so as to
meet uncertain demand situations [15], [16].
But opaque selling has its own share of demerits. Firstly, opaque selling through
intermediaries yields higher incremental revenue due to the uncertainty in demand,
this means in case of no or little demand urtcertainty, selling through the opaque
channel will faintly increase profits for the airline, when comparing to direct selling
channel [13]. Moreover, opaque channel on one side increase sales by attracting price
sensitive travelers who may otherwise not purchase, at the same time it also causes
2
reduction in sales of the transparent online SBTs of airlines and of the offline
channels such as traditional travel agencies [12]. Earlier opaque selling intermediaries
performed much better because there was little or no competition. However, today
with the boom in Online Airline Reservation Systems (OARS), the price of
discounted products may not vary much because potential buyers will be dispersed by
different opaque selling intermediaries that will be standing against each other,
striving to steal market share and forming a tacit collusion to keep prices high so as to
make profits [ 12]. This will minimize incremental revenue of a particular airline [ 17].
In fact, with growing competition in opaque selling intermediaries, product
differentiation has become difficult [ 18], so does branding and building customer
loyalty [19].
While explaining the reasons, why people prefer intermediaries over online
booking systems, researchers [20], [21] argued that earlier Business to Consumer
(B2C) systems were not flexible as they work for simple closed requests, i.e. a request
that can be direct! y mapped into formalized terms or predefined parameters, such as
dates, airports, flights etc. Furthermore, these systems could break down for more
complex requests, i.e. a request where customer is flexible with regards to attributes
such as date and destination. Therefore, Malizia and Olsen [20] have recommended an
information system between a customer and booking system to replace intermediaries.
However, the solution recommended by these researchers only covers pre-sale
flexibility issues. In other words, these systems provide flexibility only in terms of
providing general information, which could be useful in taking decision with respect
to pre-sale flexibility. So the question is if current SBTs are flexible enough? The
term flexibility here should not only be related to the booking. In case of 'e-ticketing',
problems arise when a traveler changes his/her mind or if the airline decides to make
changes with regards to times, dates, destinations, after receiving a final confirmation
of the booking. Therefore, actual replacement of human agent with a virtual
intermediate system could be attainable in a post-sale flexibility scenario, if it is really
supported in terms of 'flexibility'.
According to the Special Interest Group on Human-Computer Interaction
(SIGCHI) of the Association for Computing Machinery (ACM) "Human Computer
3
Interaction is a discipline concerned with the design, evaluation and implementation
of interactive computing systems for human use and with the study of the major
phenomena surrounding them." And within the study of HCI, human actions are
processed by computers; as a result interaction occurs between the two. This means,
humans make computer perform operations. Therefore, it is necessary to understand
such an interaction in the contexts of flexibility as well. For that reason, flexibility can
be discussed from these two different perspectives, i.e. (I) System's Flexibility
(Computer) and (2) Users' Flexibility (Humans).
1.2 Problem Statement
In designing a flexible system, it is inadequate to understand and provide only the
System Flexibility. It is equally important to also address the flexibility on the users'
side. Researcher [22) argued that even though internet is referred as a major
technological innovation of today, its success heavily depends upon assimilation of
customer expectations and preferences into the design and content of websites. In
airline industries, Users' Flexibility is more prevailing as most airlines adopted last
minute ticket selling strategy and opaque selling through which travelers can enjoy
highly discounted fares at the price of their fl·~xibility on, for instance, the traveling
dates, time or itineraries. Previous Human Computer Interaction (HCI) researches,
however, rarely addressed the users' perspectives flexibility when designing systems
interface for usability and flexibility of Online Airline Reservation Systems.
Therefore, it is important to design self-booking tools in view of customers'
preferences, expectations and online usage behavior in order to increase the Perceived
Usability of such systems.
Usability of a system can be evaluated on the basis of performance of its different
functions and from literature it is noticeable that flexibility of a system is one of the
guiding principles that provide support to achieve, develop or improve its usability.
Therefore, it is very critical for the organizations to design flexible systems that are
easy to use and can accomplish all the requirements by way of offering
customizability. And one should not overlook, that flexibility is mirrored in functional
requirements as well [23). A system is considered usable if users can accomplish their
4
tasks easily. Similarly, a system is considered functional if it otTers all the functions
required by a user to perform their tasks [ 24].
Norman's philosophy [25] says that users are good in adapting to as ystem.
However, research shows users' dissatisfaction with existing Online Airline
Reservation Systems in terms of task completion [26]. Therefore, researchers are
eager to find out ways for improving online usability of the systems, how users'
Perceived Usability of a system is formulated by its flexibility functions [27]-[31]. In
addition to techniques, methods and guidelines proposed for designing usable
systems, HCI researchers have also long argued on the importance of human factors
in designing and implementation of user-centred designs. According to Nielsen [32],
"users experience usability of a site before they have committed to using it and before
they have spent any money on potential purchases". This indicates users' Perceived
Usability in online digital environments is an important determinant for evaluating
their satisfaction in the same environment. The existing literature can be divided into
the following four research aspects where researchers currently are focusing upon to
determine usability of online systems:
• Usability Perception by Performing Content Analysis
• Usability Perception through User's Internet Adoption
• Usability Perception based on Users' Preferences and Expectations
• Usability Perception based on Online Behavior of Web Users
This research uses a blended approached and combines the above four research
areas to determine usability perception of the Online Airline Reservation Systems. It
is important because customers' usability expectation and preferences from Online
Airline Reservation Systems lacks research and empirical findings. Law and Leung
[33] had emphasized upon the need to investigate expectations of airline customers
that book their itineraries through their online self-booking tools. Moreover, the
existing evaluation of online tourism websites is performed by researchers and not by
customers. It leads to a dilemma and research gap that does not potentially address
expectations of travelers. This research therefore examines travelers' expectations,
5
preferences and online behavior (User's Flexibility) and aligns that with designing of
flexible Online Airline Reservation Systems (System's Flexibility) and users' as
evaluators of the online systems to determine its usability (Perceived Usability).
1.3 Research Aims and Objectives
This research aims to provide a framework for designing a more flexible Online
Airline Reservation Systems through investigating the associations between System's
Flexibility, Users' Flexibility and Perceived Usability of Online Airline Reservation
Systems. This was an exploratory approach and would lead to a better understanding
of the interrelationship between System's Flexibility, Users' Flexibility and Perceived
Usability of Online Airline Reservation Systems; and provide a basis for future
studies to formally develop design guidelines and/or usability metrics in the flexibility
context. To aid this aim the following research objectives are defined to address the
corresponding research questions/hypotheses:
I. To assess user needs (System's Flexibility and Users' Flexibility) associated
with Online Airline Reservation Systems.
The research questions are:
RQl: What are the issues with flexibility of Online Airline Reservation
Systems, whether or not flexibility is one of the reasons for users not using
such systems?
RQ2: To what extend flexible users can compromise with serv1ce quality
attributes of Online Airline Reservation Systems?
RQ3: How users' satisfaction with an existing SBTs is rated against their
choice of OT A feature and reflected in their integration assessment of the
same for making SBTs more flexible Online Airline Reservation Systems?
2. To propose a framework for designing more flexible Online Airline
Reservation Systems while classifying users on the basis of their Flexible
Traveling Behavior.
6
The research questions are:
RQ4: How users' perception on factors influencing Flexible Traveling
Behavior and Flexible Online Airline Reservation Systems is determined~
RQS: How to classify Users' on the basis of their Flexible Traveling Behavior
into High, Medium and Low flexible and how to investigate interrelationships
among Users' Flexibility, System's Flexibility and Perceived Usability of
existing Online Airline Reservation Systems?
3. To study the interrelationship between System's Flexibility, Users' Flexibility
and Perceived Usability of Online Airline Reservation Systems and to
determine the Perceived Usability of the existing and proposed systems.
The research questions are:
RQ6: How do service quality attributes of airlines and external variables
jointly predict flexible behavior of travelers?
RQ7: How does user Perceived Usability with the existing and the proposed
system differs?
RQ8: Is there a multivariate main effect of user's Flexible Traveling Behavior
(High, Medium and Low) on effectiveness, efficiency and satisfaction of the
proposed system?
1.4 Research Methodology
The methodology of this thesis consists of three phases which are described below:
• Phase 1: Assessing User Needs (System's Flexibility & Users' Flexibility)
Phase I was designed to achieve the I 51 research objective. The existing Online
Airline Reservation Systems were used to assess the System's Flexibility and
Users' Flexibility. In the existing Information Systems research and literature,
no study has been found to address consumer behavior on opaque selling with
7
respect to Online Airline Reservation Systems. Therefore, this phase will help
to explore the small but growing literature in designing of Online Airline
Reservation Systems by modeling upon flexible behavior of travelers.
Three pilot studies were conducted in this phase as shown below:
I. A study to investigate issues with flexibility and if flexibility is the
reason for not using Online Airline Reservation Systems.
2. A study to investigate users' flexible behavior in terms of
compromising on the service quality attributes of an airline.
3. A study to examine if integration of OT As features can make SBTs
more Flexible Online Airline Reservation Systems?
• Phase II: Classification of Users (Interrelationship Testing of Variables)
Phase 2 was designed to attain the 2"d research objective. This phase intends to
carry out an extensive relationship testing of variables and their sub-measuring
constructs so as to evolve a framework for designing of Flexible Online
Airline Reservation Systems. Two detailed studies were conducted in this
phase with the following study objectives:
I. A qualitative enquiry to explore the concept of users' perception on
factors influencing Flexible Traveling Behavior and Flexible Online
Airline Reservation Systems.
2. A study to classify Users' on the basis of their Flexible Traveling
Behavior (High, Medium, Low) and to investigate interrelationships
among System's Flexibility, Users' Flexibility and Perceived Usability
of existing Online Airline Reservation Systems.
• Phase III: Case Study (Testing the Framework)
Phase III of this study is related to a design case study and the corresponding
analysis to conquer the 3'd research objective. Participants were requested to
complete the usability evaluation of the existing and proposed interfaces.
8
Quantitative technique was used to analyze the data collected in the context of
System's Flexibility, Users' Flexibilitv and Perceived Usability of the
systems.
A detailed research methodology is presented in Chapter 3.
1.5 Scope of Research
Global Distribution Systems allow users to make reservations, from hotel booking to
car rentals, from railway reservation to e-ticketing. However, the scope of this
research is limited to Online Airline Reservation Systems only, as the research
focuses upon designing a more flexible Online Airline Reservation Systems in lieu of
users' flexible behavior. Thus, the scope of this research is further limited to the
Malaysian perspective.
Forrester research [33), estimates seventy million consumers searching online for
travel plans in July 2006, thus making online travel bookings the single largest
component of e-commerce. Different users have different needs, interests and wishes
to be served and system's effectiveness, etliciency and satisfaction may vary from
one user to another based on their usability perception. For some users a system may
be very effective but this may not be true for all. Therefore, usability of any website
cannot be improved without considering consumer intend or user behavior.
Furthermore, clear understanding of consumer intent and behavior in the case of
online airline ticket shopping and elsewhere cannot be achieved without considering
the factors that affect purchase decisions [34)-[36]. The reason is selling products
online are very different from selling in physical market and this requires a clear
understanding of online customer interest due to absence of face-to-face interaction
with customers [20). [21), [37]. Internet marketing strategies can be adjusted if cyber
marketers know what the consumers want and how they reach their decisions.
Similarly, such an understanding will help Web designers to develop sites making that
are not only popular but also flexible and effective for sales [38), [39).
The anticipation of travelers for low fares is an extremely important concern, that
airlines are faced with every day. Anticipation of travelers for low fares, gives an idea
9
about their traveling behavior, which is 'the extent to which a traveler anticipates/or
a low fare ', indicating the extent to which 'a traveler is ready to compromise on
flying conditions', and thus becoming flexible in accepting what is being offered to
them by an airline. As discussed in Chapter 2, Section 2.3.1, airlines are not in a
strategic position to offer low cost fares directly to their customers for selling left over
inventories, due to a number of potential threats to their revenue generation.
Therefore, the only business model that addresses this concern is opaque selling by
OT As, and scope of this research is narrowed down to examining the t1exible
behavior of Malaysian travelers in order to design more t1exible Online Airline
Reservation Systems that may increase the Perceived Usability of Online Airline
Reservation Systems.
1.6 Research Contributions
The first contribution of this research is the development of a framework that could be
used for further studies and to design more flexible Online Airline Reservation
Systems. The framework is a general framework that can be applied to different
reservation systems; however, this research particularly addresses the airline
reservation systems.
The second contribution 1s towards the development of users' flexibility
measuring scale. The non-availability of an absolute scale to measure t1exibility turns
the investigation into a cumbersome effort for researchers and practitioners. It is
difficult to even make any rough assumptions about the extent to which the users
would like to have additional flexibility features in online reservation systems.
Therefore, this research builds on previous and ongoing work within the disciplines of
Human Computer Interaction by introducing psychometric scales to measure users'
flexibility in terms of compromising on service quality attributes of an airline.
The third contribution of this research is within the area of Operational
Management, as it introduces a new approach of reservations to increase the
Passenger Load Factor (PLF). Under utilization of the resources, such as, air plane
capacity is one reason of low PLF and increased number of flights. Using traditional
10
airline reservation systems, airlines either cancel the flight at II th hour or send flight
with minimal profit margin, even if they receive lesser bookings in any particular
flight. One solution to address this concern is by leveraging upon travelers' flexibility.
In this research it is proposed to send flights fully occupied, which in turn could
reduce the number of flights that actually take-otT per week.
The fourth contribution of this research is to add up in the growing literature. The
concept of SBTs for disposing off their opaque inventory directly has not been
adequately considered in Information System Research and Literature. The concept
requires extensive research especially in academic discipline [ 40] and as highlighted
by Jerath et a/. [ 41] a number of studies have focused upon airline revenue
management systems, however attempts to empirically verify those findings are a
few. Furthermore, the existing opaque selling literature lies at the intersection of
consumer behavior and revenue management operational strategies [ 40). However no
study has been found to address consumer behavior on opaque selling with respect to
Online Airline Reservation Systems as most recent papers as well as researches fall
within the marketing domain [ 40). This research therefore contributes to the small but
growing literature in designing of Online Airline Reservation Systems by modelling
upon flexible behavior of travelers.
The fifth contribution of this research is the IS theory, based on empirical findings
and analysis in support of the proposed framework for Flexible Online Airline
Reservation Systems. Theory building from case studies is considered to produce
novel theory, and is testable with constructs that can be readily measured and
hypothesis that can be proven false.
The sixth contribution of this research is the development of a prototype in terms
of proposed Graphical User Interfaces (GUI) to handle flexible and inflexible
travelers differently. If a system provides ranges of dates as flying and source
destination options at different fares, flexibility of the system is enhanced in its
Perceived Usability in the eyes ofthe ±lexible travelers.
Finally, this research provides empirical results of the real case studies on the
existing and proposed system.
II
1. 7 Organization of the Thesis
This dissertation is divided into the following 6 chapters:
Chapter I provides a research overview. It gives the research background while
defining the problem statement, research aim and objectives, research methodology,
research scope and contribution. Finally, it outlines the overall chapters of the
dissertation.
Chapter 2 presents the literature review. This chapter introduces flexibility of
Online Airline Reservation Systems from two different perspectives (i.e. System's
Flexibility and Users' Flexibility) and also provides a conceptual linking between the
two. Furthermore, different aspects of Perceived Usability have been discussed in this
chapter in order to provide a conceptual linking between the usability of Online
Airline Reservation Systems and Flexible Traveling Behavior ofthe travelers.
Chapter 3 describes the research methodology of the thesis. The overall
methodology is divided into three phases. Phase I addresses user needs (System's
Flexibility and Users' Flexibility), Phase II gives classification of users and
interrelationship testing of variables and, Phase III provides a case study.
Chapter 4 reports the statistical analysis of the research. It includes results of
different pilot studies and the case study. It follows the research questions to organize
the results obtained through corresponding hypotheses.
Chapter 5 presents the discussion of the dissertation. This chapter follows the
same pattern by discussing and elaborating facts related to individual research
questions reported in Chapter 4. Moreover, recommendations for the proposed
Flexible Online Airline Reservation Systems are also discussed in this chapter.
Chapter 6 summarizes the thesis. It highlights the research and emphasizes on the
importance of the proposed framework. Moreover, major findings of the work and
recommended directions for the future work an: also presented in this chapter.
12
2.0 Chapter Overview
CHAPTER2
LITERATURE REVIEW
In this chapter System's Flexibility, Users' Flexibility and Perceived Usability of the
Online Airline Reservation Systems is presented. Section 2.1 is dedicated to explore
flexibility from two different perspectives (i.e. System's Flexibility and Users'
Flexibility) and to find out the conceptual link between the two. Section 2.2 covers
different aspects of usability perception and provides a conceptual link between the
usability of Online Airline Reservation Systems and Flexible Traveling Behavior of
the travelers. Section 2.3 is devoted to explore the relationship between flexibility and
usability and also to study the role of opaque selling in System's Flexibility and
User's Flexibility. Section 2.3 presents the summary of the chapter.
2.1 Flexibility and Online Airline Reservation Systems
Flexibility of a system is one of the guiding principles that provide support to achieve,
develop or improve its usability. It is very critical for the organizations to design
flexible systems that are easy to use and can accomplish all the requirements by way
of offering customizability. Given the importance of internet shopping as a source of
income for the airlines and high user demand, in-depth research is required.
2.1.1 Flexibility Concepts
The notion of flexibility has been addressed in many disciplines and from many
different perspectives. The oxford university's dictionary on Business and
Management defines flexibility as "the ability to adapt an operating system to respond
to changes in the environment" [42]. In case of manufacturing, one may distinguish
eleven different classes of flexibility: machine, material handling, operation, process,
product, routing, volume, expansion, program, production and market flexibility [43]
(45]. In the discipline of systems engineering, the flexibility of a system is understood
as "the ability to respond to change" [ 46]. Product design literature defines flexibility
"as the ability of companies to frequently upgrade their products to meet the rapidly
changing technologies" [47], (48]. Each of the above definition defines flexibility in a
different perspective, but the fundamental meaning of this term remains consistent
across all definitions which: "able to flex."
As defined by the Special Interest Group on Human-Computer Interaction
(SIGCHI) of the Association for Computing V!achinery (ACM) "Human Computer
Interaction is a discipline concerned with the design, evaluation and implementation
of interactive computing systems for human use and with the study of the major
phenomena surrounding them." In HCI, human actions are processed by computers;
as a result interaction occurs between the two. This shows that humans make
computer perform operations, therefore, it is very important to understand human
computer interaction in the context of flexibility as well. Hence, flexibility can be
discussed from these two different perspectives, i.e. (I) System's Flexibility
(Computer) and (2) Users' Flexibility (Humans) as shown in Figure 2.1.
System's Flexibility
I User's FlexibilitY]
Figure 2.1: Two Different Perspectives of Flexibility
2.1.1.1 System's Flexibility
Within the HCI discipline, System's Flexibility is referred to its ability to respond to
internal or external changes. However, th<: ambiguous characteristic of word
"flexibility" [46] has forced authors to explain flexibility differently as shown in
Table 2.1.
14
Human computer interaction studies are conducted to develop or improve the
safety, utility, etlectiveness, efficiency. usability, appeal of the systems that include
computers as shown in Figure 2.2. Of which, Usability of the systems is described by
researchers as a "measure of the ease with which a system can be learned and used, its
safety, effectiveness and efficiency, and attitude of its users towards it" [49]. While,
ISO defines Usability as "the extent to which a product can be used by specified users
to achieve specified goals with effectiveness, etliciency and satisfaction in a specified
context of use" (ISO 9241-11 ). The principles that provide support to achieve,
develop or improve Usability of the system include, (I) Learnability, which is the
ease with which users can use the system effectively. (2) Robustness, which is the
level of support provided to the user to achieve its goals and (3) Flexibility, which is
basically multiplicity of ways the user and the system exchanges information.
Table 2.1: Definitions of Flexibility in the Context of System Engineering
Author
Nilchiani
Saleh
Ross
Definition of Flexibility We define flexibility as the ability of a system to respond to potential internal or external changes affecting its value delivery, in a timely and cost-effective manner. Thus, flexibility is the ease with which the system can respond to uncertainty in a manner to sustain or increase its value delivery. It should be noted that uncertainty is a key element in the definition of flexibility. Uncertainty can create both risks and opportunities in a system, and it is with the existence of uncertainty that flexibility becomes valuable. Flexibility should be sought when: I) the uncertainty in a system's environment such that there is a need to mitigate market risks, in the case of a commercial venture, and reduce a design's exposure to uncertainty in its environment, 2) the system's technology base evolves on a time scale considerably shorter than the system's design lifetime, thus requiring a solution for mitigating risks associated with technology obsolescence. The only difference between flexibility and adaptability is the location of the change agent with respect to the system boundary: inside (adaptable) or outside (flexible). Of course the system boundary could be redefined, changing a flexible change into an adaptable one, or vice versa. The fungible nature of the definition is often reflected in colloquial usage and sometimes results in confusion. If the system boundary and location of change agent are welldefined, confusion will be minimized.
15
Reference
[50]
[51]
[52]
0.'\.'\.~ HCI Constructs
Y-'-'~~~~"..~~~~~~" Safety Utility
Figure 2.2: Different Elements of Human Computer Interaction
Of the three principles, Flexibility is related to taking input/output in different
forms and examined with respect to (I) Dialogue initiatives, (2) Multi-threading, (3)
Task migratability, (4) Substitutivity and (5) C:ustomizability as shown in Figure 2.3.
Customizability refers to adaptability of interfaces to suit different needs, and it is
achieved by way of (i) adaptability, where users can adapt the user interface, (ii)
adaptivity, where the user interface can be adapted by the system and (iii)
personalization, where the user interface is tailored towards the individual user. While
system driven interaction hinders its flexibility, user-driven interaction is considered
to be strongly favourable.
Dialogue Initiative
Multi Threading
I Leamability l
Task Migratability ~ Flexibility
I ~ Usability
Adaptability Substitutivity I
Robustnous J Adaptivity Customizability
Personalization
Figure 2.3: Flexibility as the Multiplicity of Ways for Information Exchange
2.1.1.2 Users' Flexibility
The User's Flexibility is nothing but users' ability to rapidly change from one course
of action to another, i.e. "flexible behavior", and it is referred as a hallmark of human
16
cognition system [53]. Webster's Dictionary defines cognition as "the act or process
of knowing in the broadest sense; specifically. an intellectual process by which
knowledge is gained from perception or ideas". Empirical research into cognition is
usually scientific and quantitative, and involves formation of mental models to
describe or explain certain behaviors. In context of t1cxiblc behavior of users, human
cognition system may thus be examined from the perspective of cognitive
psychology. As mentioned earlier, Human-Computer Interaction (HCI) research is
intended to explain interaction between humans and the computer technology. And in
order to provide a scientific explanation to human behavior (e.g. user interface design,
information visualization, etc) many principles, theories and concepts from cognitive
psychology are deployed in HCI [54]-[60] such as Perception, Categorization,
Memory, Knowledge Representation, Language and Thinking as shown in Figure 2.4.
L Perception
Categorization
Memory
Co~n-;;~
l Psychology 1
Logic I j I L::::e_ __ j
,.I Thinking --~
I_ Decis~~~aking
Problem Solving
Figure 2.4: Building Construct of Cognitive Psychology
Thinking refers to any intellectual or mental activity resulting m ideas or
arrangements of ideas and within the context of HCI. thinking simulates human
behavior, which is eventually translated as an action taken on part of users, in the
form of making choices, performing logical operations. formation of concepts,
problem solving and decision making [61]. For understanding t1exible human
17
behavior, decision making is an important reflection of users thought process.
Decision making is a mental process which results in selection of a course of action
among several alternatives. At the end of every decision making process, an output is
produced in the form of a final choice or selection, which can be in the form an
action, or an opinion of choice. Decision making process is an active research area
since it examines decisions of users in context of their unique set of needs and
preferences, therefore, reflection of users' Hexible behavior can be seen in the
decisions they make [62], [63].
2.1.1.3 Conceptual Linking between System's Flexibility and Users' Flexibility
In order to understand users' Hexible behavior, it is first important to understand the
contexts that govern users' behavior towards being flexible. In case of System's
Flexibility, users interact with computer systems in order to accomplish tasks. While
the System's Flexibility is reHected in its customizability features, therefore,
developing an understanding of system's customizability in terms of affecting users
Hexible behavior, requires a science base in the form of systematic knowledge of what
governs user's Hexible behavior' and inHuencing upon their decision making process
as shown in Figure 2.5. Thus, three variables have been identified in this basic
conceptual framework: (i) System's Flexibility, (ii) Users' Flexibility and (iii)
System's Usability.
From literature review, it is found out that users' flexible behavior is reflected in
their decision making process, while Syst,~m's Flexibility is translated in its
customizability features.
I. System's Flexibility has a linear relationship with User's Flexibility due to the
following assumptions:
• A Hexible system (customizable) can reinforce users' Hexible behavior by
inHuencing upon their decision-making, even if they were inHexible or
partially Hexible initially [25].
18
• On the contrary, if a user is flexible with respect to making decisions, it
cannot still reinforce System's Flexibility through customizability, even if it
was inf1exible or partially flexible, initially.
2. System ·s Flexibility is one of the principles that provide support to achieve,
develop or improve usability of the system. System's flexibility thus has a
linear relationship with System's Usability.
3. Users' flexible behavior in terms of their decision making influence upon the
usability of a system. Thus Users' Flexibility has a linear relationship with
System's Usability.
Adaptability
~-·--:~~~:~ity l ________ _
Personalization
Dialogue lmtiattve
Multi Threading
Task Mtgratabiltty
Substitutivity
{ Customuobility
)1- System's Flexibility
!
Users' Flexibility
Cognltivo P•yc~ology (~ht~~;ng) J A
Chol<e . -] i Logic J [con<ept Fonn,.ion : D«l•ion Making 1 Problem Solving
Figure 2.5: Conceptual Linking between System's Flexibility and Users' Flexibility
From literature it is concluded that flexibility can be discussed from two different
aspects, one from System's perspective (Computers), and second from User's
perspective (Human). System's Flexibility translates into its customizability in
achieving the defined usability objectives of the system which are effectiveness,
efficiency and satisfaction.
Likewise, in context of User's Flexibility, this section concludes that different
users may have different needs, interests and wishes to be served and system's
19
effectiveness, efficiency and satisfaction may vary from one user to another based on
their usability perception. For some users a sys1tem may be very effective but this may
not be true for all. This drives the need of integrating cognitive ergonomics into the
framework, to understand Users' Flexibility in designing of systems. Moreover, from
literature it is found that Users' Flexibility is ret1ected in their decision making
behavior. Further elaboration and validation of the conceptual linking between
System's Flexibility, Users' Flexibility and Usability can be found in the
methodology chapter Section 3 .2.1.
2.1.2 Airline Reservation Systems
Airline Reservation Systems (ARS) keep record of airline schedules, fare tariffs and
passenger reservations. ARS are developed to enable productive and effective flight
reservations for an airline. ARS eventually evolved into the Computer Reservation
System.
2.1.2.1 Computer Reservation Systems
Computer Reservation Systems (CRSs) are the computerized systems used for storing
and retrieving information such as, airline reservation systems, car rental systems and,
hotel reservation systems [I].
CRSs became increasingly popular C."~ • .:> their immense potential in handling of
reservations and companies could foresee an increase in their yield matrices.
However, CRSs offer advantages but the strength of their positivity depends upon
how well and at what level systems have been integrated [ 64].
CRSs are equipped with enhanced functionalities and features that provide
companies with an integrated one stop solution to manage sales, customer relationship
management, marketing plans, resource planning and personalized customer care and
attention. With these features, CRSs help in processing reservations and at the same
time support decision making processes.
20
The development of CRSs started at the beginning of the 60s when the first
electronic travel hooking was launched by Semi-Automated Business Research
Environment (SABRE). Subsequently. American Airlines and IBM joined hands with
SABRE to launch first of many airlines owned and operated CRSs.
Initially, these systems were used at airlines· basic and internal reservation
centres, but its true potential was realized quickly and travel agencies grabbed the
opportunity for its deployment. It helped travel agencies immensely in terms of
providing accurate schedules to travelers, fares. instant information on availability of
seats and extended efficiency internally as well with respect to strengthening the
distribution channels. However, CRSs came with their share of drawbacks as well.
CRSs did help in reducing the costs of travel agencies when compared to manual
reservation systems that were based on telephone confirmation and checking [ 65],
however, they were still criticized as "inflexible dinosaurs" since they were not
adaptable enough to meet growing business demands, that requires robustness with
regards to offering services and additional features for reducing high distribution costs
in a more flexible manner [I]. Airlines, being the true originator of CRSs enjoyed
more competitive edge than problems that had arisen due to inflexibility of CRSs.
Especially when comparing with travel agencies. airlines were in a more control
situation with respect to scheduling of flights and could even influence upon market
share. For example, in 1985, U.S. travel agency sales had risen to $54 billion, which
was more than nine in I 0 agencies with sales greater than $1 million and had
deployed CRSs. Travel Agency revenue had surged 400% over the same period, while
agency employment increased by only 20% [9].
In early nineties, with the consortium of four large CRSs compames. Global
Distribution Systems (GDSs) had emerged into the scene. It is important to mention
here that CRSs are not to be confused with GDSs since they are electronically
connected to one another. CRSs run on mainframes. minicomputers or
microcomputers and are connected through data communication links to terminals
within various branches of the company for bookings. On the other hand, GDSs are
the systems that book and sell tickets for multiple airlines and use internet gateways to
allow users for making reservations, from hotel booking to car rentals, from railway
reservation to e-ticketing as shown in Figure 2.6. The emergence of GDSs and their
21
connectivity with CRSs has brought hundreds of thousands of travel agents and other
distributors with thousands of suppliers on one single platform, hence resulting in
improved efficiency, facilitating control and rapid response time to both customers
and management (3 ]-[ 5 J. This is particularly true in case of global tourism industry
since it heavily deploys CRSs to process their reservations through GDSs to perform
basic functions of reservation process, such as product presentation, reservation, fare
quote & ticketing and additional services [!]. This is reflective in tourism or
hospitality industry where over the years, electronic reservation systems have
provided greater operational benefits in temts of yield management, e-marketing
strategies as well as productivity benefits [64]. Likewise travel agencies through
GDSs enjoy the freedom to make reservations directly from their terminal with any
airline, on any continent. This saves much of their coordination time and effort that is
required in settling negotiations.
Figure 2.6: Distribution Channels for Airline Reservation Systems
As mentioned earlier, GDSs has brought hundreds of thousands of travel agents
and distributors from different countries and continents in nexus, it has thus acquired
22
the status of the nervous system in global reservation system. The GDSs generated
more than $9.6 billion in revenue and more than 1.1 billion transactions in 2008, just
over 2, I 00 transactions per minute. The infrastructure of each GDS can support
volumes far greater than this [9] since airlines and other distributors that are in nexus
with GDSs and provide access to schedule and fare to travel agencies - both offline
and online [9].
2.1.2.2 Self~Booking Tools vs. Online Travel Agencies
A travel agency, also called a travel bureau, is defined as business that attends to the
details of transportation, itinerary, and accommodations for travelers [2]. Travel
agency acts as an agent, just like a retail storefront that books and sells tickets on
behalf of many airlines. Traditional travel agencies hold a large portion of travel
booking industry, due to a number of factors such as. face to face service to
customers, provision of personalized services and realistic solutions for providing
reservation arrangements, comfort in country of destination, and special packages or
promotional deals. On the negative front, traditional travel agencies are blamed for
practicing restrictive practices such as racking whereby they promote traveling
brochures of those companies only who pay them highest commission. The traveler is
unaware of possible alternative options and considers them to be the only once or best
option available.
On the contrary, an Online Travel Agencies (OTAs) operate through a travel
website on the World Wide Web, dedicated to providing updated travel related
information, guidance and travel reviews [2 J. The travelers interact with the virtual
interface of the online travel agency which allows them to search and book their travel
plans. The online reservation process does not involve personalized attention on
behalf of online travel agency and still this does not seem to be a matter of concern
for travelers. According to Forrester research [33], approximately seventy million
consumers searched for travel plans online in July 2006. thus making online travel
bookings the single largest component of e-commcrce. This also makes online travel
agencies an important part of the overall equation for flexible reservation systems for
airlines. The recent growing acceptance of online travel agencies is credited to their
23
meta-search engines feature that provides fare: aggregators to travelers. Meta-search
engine as the name indicates conducts search across multiple independent search
engines and gets live availability of flights through "screen scraping" process, which
crawls through the airline websites and extracts content by way of human-readable
HTML feed. The content extracted from various airlines website is then displayed to
the users in the form of fare aggregation, i.e. all results on one screen. According to
PhoCusWright Report 2009 [9], the overall share of online travel agencies in US
travel market alone was 13% in 2008 and projected to touch 16% in 2011 as shown in
Figure 2.7. On the contrary, the share of conventional traveling agency was 33% in
2008 and is projected to suffer a decline 3% in 2011. This is further justified from
Yahoo Travels' claim which says that 76% of all online travel purchases occur as a
result of search function. Jupiter Research in its Travel Consumer Survey published in
2004 pointed out that "nearly two in five online: travel consumers say they believe that
no one site has the lowest rates or fares." This therefore created a niche research
dimension for OTAs to look at different ways for integrating additional features into
their reservation system so as to optimize aggregate travel search and provide lowest
rates from multiple travel sites, to eliminate travelers' verification need from site to
site.
It is also interesting to note from the chart as shown in Figure 2.7 that supplier
branded web sites also will experience an estimated increase of 3% in 20 II. Suppliers
branded websites are Self-Booking Tools (SBTs) providing direct linkage of the
passenger with the airline industry. They provide carrier-direct bookings facility to
travelers, without having them going through the hassle of other intermediaries. These
booking are just like going to the reservation office of a specific airline physically and
are popular among travelers who remain loyal to their favourite brands of airlines and
prefer to travel only through them. Another strong reason for travelers to opt for SBTs
is their ability to earn flying rewards, which ultimately makes them more loyal
towards a particular brand of airline. Likewise, in order to differentiate their
reservation channel from others, airlines have started to invest heavily in their online
SBTs capabilities, offering more features and convenience for travelers such as
tracking, managing and redeeming air miles. Moreover, airlines also invest massively
on branding of their image and securing loyalty of customers by offering reward
24
mileage bonuses. In addition to this some airlines have gone to the extent of imposing
fees on GDS bookings for their carrier. For example Lufthansa airline in its Preferred
Fares Program launched in 2008, imposed fees of €4.90 per ticket for travel agencies
in Austria, Germany and Switzerland that made reservations through GDSs [6].
2008
• Suppl•er Brmclo!d W<'b Sir"'
• Ortlif'f! Tt•~fl A.J.-roet•!S
• fr•l'el AgenkM/ TMC~·
• Coo~DCf!n'"'' '.'.'~lk· n
2011
Figure 2.7: U.S. Travel Market by Channel, 2008 and 2011 (Projected)
Source: PhoCusWright U.S. Online Travel Overview Ninth Edition
On the contrary, OT As cash upon the nexus of CRSs and GDSs and act as a
central hub for price differentiation and comparison. They provide discounted fares
and 24 hours service. Their fixed costs are lowest, since there is no requirement as
such to set up physical offices with state-of-the-art facilities at prime locations. Their
success is derived by innovational strategies, as a result they hold lion's share, 50%
(average 2006, 2007 & 2008) in the travel industry.
A comparison chart on SBTs and OT As is presented and discussed as shown in
Table 2.2. The table presents innovative attributes and function that have contributed
immensely towards the popular acceptance of OTAs over Airlines' SBTs and are also
widely common among travel companies in recent years (15], [16]. Functions such as
product presentation, reservation, quoting & ticketing, post-sale services, low fare
notification, dynamic packaging and flexible alternative date search are also
performed by SBTs, however, OT As get an edge over SBTs in terms of providing
matrix display, opaque fares, alternative airport search and hotel search.
25
Table 2.2: Comparison Chart on Offered Features by SBTs and OTAs [1], [9]
Feature
Product presentation
Reservations
Quoting & ticketing Additional services
Matrix display
Alternative airport search
Hotel search, results display & sorting
Opaque fares
Low-fare notifications Flexible & alternative date search
Dynamic packaging
Description
is the presentation of services or products in all aspects of travel industry. is used for making reservations for the offered services and products. relates to providing fare quotes and generate receipts for the given services and products. post-sale features and user prompting for their guidance throughout the reservation process . using this feature, users may click on any cell within the matrix to sort airfare search results by price, airline and number of stops. It was initially introduced by Orbitz but these days it has become a standard for all OTAs. allow travelers to search across multiple departure and arrival airports so as to find the lowest possible fare or most convenient schedule. this feature allow travelers to display, sort and compare options from hundreds of possible hotel search results. • address or landmark search and sorting • map-based search results display • traveler reviews included with the results • multiple sorting options, including price, star
rating, brand, guest rating and amenities were initiated by Priceline's Name Your Own Price airfare bidding model. In this feature users are offered heavily discounted tickets with not specified time or flight number. They are usable at the discretion of the airline. a feature in which customers are intimated via email to opt for specific promotional deals. allow users to search and compare flight options across multiple departures and return dates so as to find the lowest possible fare. initially made famous by Expedia where users are allowed to shop for multiple components m a single search, such as "Flight and Hotel".
SBTs OTAs
Matrix Display - Orbitz is the pioneer in OTAs who initiated the concept of
matrix display. This feature allows users to click on any particular airline offered fare
to see the further details such as departure and arrival timings. As shown in Figure
2.8, there are number of airfares from New York to Los Angles offered by multiple
carriers on the specified dates e.g., departure: 29 December 201 0; arrival: 29 January
26
2011 . 1n the matrix, carriers are organized in multiple columns, stops in multiple rows
and the airfares are placed against airlines and stops.
rRBITZ .t ;...."1'"'t ·c J't."J "" I ~t '"
-"bl!U I I'L~I ~ I ,.,. tl • ··~
(!JlJilfi10D
10 FU;Mr IOOliNHUS +PRICEWU,AII«
ti'tt~l.t
OLUntiiOII
Flldtli&)nlly A~ ~ - ,,,uf"Jple AL l!!.!lu Jlj ~
s.... Plica C1rritr1 " ~ ~ 6:mro l!!!l!!ll
Uon·t!op ~71 $527 $665 $884 $655 :::..1 t:!ll ....... ~~~ 1!. ::;.a-1 .1&.&.&.6. ~:a ' 2It
~57 $527 $530 $521 $556 ~ :.:at ::·•I L_ ,.,,_ 0 Jol till
l!...1!m S530
:::sl .,...
•.-n•·eet·wsenn~Soolots..-gt«lt!> ·ow!Wtr-•.,....-..a!l.llu !.;;H"tl tt;;•;t c"'t rau !I! a;st
l!ll!:tll
$711 sn• :-:·.11 ........ , .. ,~
S72• ::ao• .:!.,.
S:Cl!t (.fltrsrei '121lrt ~ WChan~.C~IIeltlf06. CW9ftitltlr~D ... M,fi.'ltD~
~ ~. ... !JI.W ~ l!.2!!W ~ ~ ~
$751 $757 ·:~•I L :::•1
$1.112 S838 l ca•l:£
[-] FHCtacll
Figure 2.8: Lowest Airfare Search Results through Matrix Display [66]
Alternate Airport Search - Alternate airport search feature helps users to find
the lowest possible fare or most convenient schedule across multiple departure and
arrival airports. Using this feature as shown in Figure 2.9, one can click on the
checkbox " include nearby airports" to see the search results on the specified as well as
nearby airports. Orbitz provides the flexibility of choosing "include nearby airports"
for source and destination airports.
Lowesl price Departure ume
Change Search s•c .... •ow..-s\ pnu:d ~'bghta • .:.~ ,. ... ..s c.: :r l!t -=~
from :ty·.~-,..,.. r---.- 0
S530 -su :..u' . .., • s:s2Uso .... c•'''" NO BOOKINHEES Asf§~~~CE • C)
"""CI.'"U~I~ ~
To :f)·•-•= .. ~ __ \
lea\lt
lU~-1·
I :eco1 •
r:::;: .... . :e:n t;IOpm .~om. e 3:00pm
Virom Amenc. 928
Sa'" =ra .. e&c.e c_,. 1 Lcs ...,,;en : .. .,Aj.
fJtlhl tArr11!1U
: .. .,, =-t•.tJ:: I'J "t :U:.-a:J:·t =:· .. .~-~:r.: G.,:-, :~a , ===-· •.:r
1U -1Prt:•,,.;-
fry ro LA. trvm \.tf Lot Anpltt =-t'M _.;r =--:e:•: :::.. •.: t: -::.. 1- ;"" t :u: :uk !y.:, .,, c,;nloJ:AnctcJes.
tl!!lll!1
AlfltCe uc-ets )Os angeles
=· """ · ···:1: :-t:.=•-f.::J:)l3
Figure 2.9: Alternative Airport Search to Find the Lowest Possible Fare [66]
Hotel Search - OT As offer hotel search feature that provides the option to
display, sort and compare hundreds of possible hotel search results. As shown m
Figure 2. 1 0, this feature allows address or landmark search and sorting, map-based
27
search results display, traveler reviews included with the results and multiple sorting
options, including lowest price, distance, star rating, brand and amenities.
l os Angeles lntemauonal, los Angeles. Calt!orma. Unned S:ates I Check-in:
;, t~ . :ec 2i <C 1 ~ I Check-out: Sa~ .a" 1: '" 11 l!itghts: 1-1 Room(s): 1 Guest(s): 1
Why book on Orbitz? Price Assurance + Low Price Guarantee + tlo Orbitz hotel chan e or .. an....,.......,..._____..o
Map large llli!JJ 327 matching Los Angeles International hotels C)
Sherman Burbank~ E BestValues .;~e··"";t ~ ~ -~ Oaks ~~~~~ Pa
D
Wes111
Custom Hotel LAX. A Joie de Vivre Boutique HoteL <=>
*** c:::> 30', off a t Chic Boutique Airpor1 Hotel S:ay'" sr; e t ) ~ \. ~ ~ eas; or .. e :o Ven ce BeaCI' an~ s~urt-e ser> ce •rc,., l'-\. F·ee .re ess •ternel l'cre
Four Points b Sheraton Los An International Airport *** ~~I R~·le¥, -c:::::>·c::::~...._ _ _.._ _ __.,
Figure 2.10: Hotel Search, Results Display & Sorting to Compare Options from
Hundreds of Possible Hotels [66]
l:l.t!.:
Opaque Fares - Opaque selling intermediaries have become an established
distribution channel for the travel industry [17]. This form of opaque selling carne into
limelight in 1998 when priceline.com's, Name-Your-Own-Price emerged with an
opaque selling business model where both the itinerary information and the identity of
the airline carrier were hidden from the traveler, until the bid was purchased as shown
in Figure 2.11. Next, major U.S airlines established Hotwire to compete in the opaque
segment of intermediaries. However, it carne up with a different opaque selling
business model, which was not based on bidding mechanism but rather posting a price
for an offer that concealed key itinerary information and airline identity as shown in
Figure 2.12 and 2.13. Since then many opaque selling intermediaries have appeared in
the international travel market, such as cheaptickets.com, onetravel.com etc. and
popularized this selling mechanism [67], [68]. However, the common aspect of all
such opaque selling intermediaries is that they are based on hiding descriptive
attributes of the service to be provided; as a result travelers cannot full y predict the
ultimate service provider or the airline.
28
nee· and Save up to 40% on Fllghts1
lle r. Yor:< Crt) IIY tc Los Angeles CA Wed Dec 29 to Sat Jan 29- 1 t1cr.et "
Select D :> '1'Jre and Arrival Airports Please note that r la.,e Your O·.•n Pnce fl1ghts are eccno"'ly (coach :lass on I)
Departure Airports
::11 Je"' York C1ty - John F Kennedy Inti IIY ( K}
llew Yor~ Ctty - La Guardta IIY tl I Je,,curgh - Ste.~art Inti I JY 1. 1
Burtan~ Glendale Pasadena CA '· L Long Beach- Daugherty Field CA ( ,sr
Whrte Plams- 'Nestchester County IIY (l-IP',) Santa Ana- John' ayne ' Orange County CA iSrJA) lsli:l - Long Island lv1acArthur I JY ( )
Name Your Own Price
Passenger and Ticket InforMation Your tnp \'rill start oeh•een 6a"" and 10om on your tra-.el dates Although "'e ah1ays Icc"< for non-stoo fl1ghts first Pncellne fl1ghts rra~ ma~e uo to ons conne:tion each ,,ay Your exact flights and llrres y,ill ce showl1 to you once your purchase IS
complete We .\~ II issue con>ement electromc tiC(ets
Please enter e'l(actly as they apoear on a drr.ers license or other offictal photo ID Passengers ut de ·r• , ~ must be accor1pamed :Jy an adult Note al and s~at t e'e e.11 are not guaranteed
If )'OU already hcr.e a pocehne orofile and ould h~e to access your stored fltght pref;~rences tc s g'
Adult Passenger 1: Under 18?
First tlame IA1ddle IJame Last Name: Suffix
Please PfO\lde ~our TSA Secure Flight 1nformat1on noW" ~eam ore Date of Btrth Gender ~-'«:- • :a;· • v~- • l.lale Female
Infants: I luf"'oer of unllc~eted passenQers
Figure 2.11 : Opaque Fares Offered by Priceline for Discounted Tickets without
Specifying Carrier, Time and the Route [69]
29
H~twire· ''"
t-Int U"'O ,. f4
h d I r r I>
O;.; t ""tt )'C"'"~tt~::!!• t"V:,. a"C • ff"t f ~I"COW;,;~:O~CC:a;, t •~ot • .t. t :111 St t "C" ·#~' • !"t 1 u: fares ct: .· ttr :rete aa·t• wr.l'! C:""t cac:.~ "' e.1 eel
Savmgs =-v: l""l(:~t .-.e"'tl t a;:, "''O"' 'c'" '~ "';" ~;"'" :~"'(
' ~e .. enc 3-1! nights •·6 nights !·7 ll!Q'I.t~ 5·B nlghl! 7·9 nights B·W nl.lltS 1C>-12 nights 1 ~·1 .: n1 nt•
g ,.
Travelln~mellm" t>etweeo ("12T~1~ t~C ~-,-~ - 1 -.u
Pt • .,.U! ....... w ...
... "' u .. .,.
. .......,. .... ._... .. , .... ........ ~ ..... Cho1ce sgg · ~..
Figure 2.12: Opaque Fares Offered by Hotwire for Discounted Tickets with Specified Time or Flight Number [70]
Your Flox1ble Dato Suarch results 1 ~·1S nichttn~' from r' '" 'lor, . C•tf uv c LC'\~ to Loa -noel•• C'- JL!..!..:. ' "ltd Dec 2Q to Thu J :.n :7 • 1 ttc· et
Sun 1.1on
~ l!!.!
IC ~
Tue
= " ~
•nd
$701* Dep..n: · .• ·.: Ctc li. 2C1C Retum: o\ t~ · ' "' 12 2C 11
14-nlght u tpa Wed
Otc 2S = .lli.!
12 1ill
'C"'":i:t-; ~· CA,_:: .. '"cl !•xes a 'I :::I •tet +g; .. t•·&'U •II
15..ntohc trtpa
Sun r.1on Tue Wed
C•c~
hll
1ll.!. !..!.t..! lli!
Thu
.?C l!.!.!
I! 1.0.!
Thu
!~
~
Frl
!l .u:.:.
U!l
••• .... 1
i!!!
uu
- r-o cate•
11t!..of"" =~~'
"'•ten Pte &:.<~• P~'•ol'\ ~ t.l$ 114 •a I'll I'IC\Ioell II "1V4'1 •lldt~n ,.,.c•rtot ;::,o.:J .~•ec:t 'Mj t.o01lcl2
1•-ftJQhtlnpdfltXlrflnQ.. JilllS. :;011 £1
$292* ro .. "':::l~ ... ~ rcr ~·f'So .. .. , t.••••"~'u•
... Sol
" .. 1 " I ll!..:: i:±J.
!.Z.!..l !..!l,;
O.pan::S•t·•"'~ 2:11 Retum: S.1 Ja'" 22 2011
' Chooso 1hLS folm 0 ____ ]
Figure 2.13: Flexible Date Search Results on the of 14-16 Nights' Trip Length [70]
In this section four OT A features have been identified and discussed that are not
integrated into SBTs of the airlines mostly because of the practical implication of
each OT A feature, which is not very feasible for the SBTs to opt for. However, it is
essential in this research to understand if the integration of same OT As features could
make SBTs flexible OARS, and if the answer was yes then how come that could be
done. Therefore, preliminary but comprehensive researches need to be conducted with
airline executives to examine their subjective satisfaction with existing SBTs by the
airlines.
30
Furthermore, it is also important to explore the recommendations for making
SBTs flexible Online Airline Reservation Systems in order to increase the usability of
online reservation systems. Further elaboration of the preliminary research can be
found in the methodology chapter Section 3.2.3.
2.2 Usability of Online Airline Reservation Systems
Researchers argued that, the design of a website determines online purchase decisions
and revisit intentions (71]-(73]. The essence of quality for a successful website has
been addressed by many authors time-to-time [74]-(78]. The quality of a website is
referred to its usability and functionality (24], [79]. A website will be considered as
useable if users can accomplish their tasks easily. Similarly, a website will be
considered as fi.mctional if it offers all the functions required by a user to perform
their tasks [24]. Thus quality of a website can be evaluated on the basis of different
functions they offer and the performance of those functions. Therefore, it is very
critical for the organizations to design a website that is easy to use and accomplish all
the requirements.
In the last twenty years different lines of research have focused on identifying
certain factors influencing acceptance of information systems and have provided
models and theoretical proposals. Social Cognitive Theory [80], Diffusion of
Innovation Theory [81], the Theory of Reasoned Action (TRA) I Theory of Planned
Behavior (TPB) [82]-[84], the Triandis Model [85], Human Computer Interaction
research [86]. the Technology Transition Model [87], and Social Network Theory
[88] are representative examples.
In particular, Technology Acceptance Model (TAM), as introduced by Davis [79]
where external variables have been identified as factors that tend to influence upon
systems perceived ease of use and perceived usefulness. In other words, the model
explains how users accept and use technology on the basis of perceived usefulness
and perceived ease-of-use of the system influenced by external factors. The contextual
interpretation of any event is determined by contextual factors that reinforce viewers'
schemas, formulate characteristics of the surrounding environment and ensure
31
effective collaboration between the two. In case of news processing, for example, the
contextual factors that reinforce viewers' schemas are their lifestyle, political
socialization, prior knowledge and life experiences, current needs for various types of
information, and attitudinal factors such as interest in news and perceived credibility
of sources [89]. During the past two decades, TAM is considered the most prominent,
powerful and parsimonious theory for describing an individual's acceptance of
information systems (90]-[93]. Many other researchers have also contributed to the
list of external variables [94]-[96] since original TAM model has more than seven
hundred citations to its credit and has been adapted and extended in many ways to
date.
TAM and TPB, have received considerable attention from the scientific
community and its use has been extended to the study of tourism services [63].
However, TAM and TPB have successfully explained behavioral intentions, previous
research pointed out that TAM and TPB' s fundamental constructs do not reflect the
specific influences of usage-context factors that may alter users acceptance. Usage
context factors are based on users' contextual interpretations that are based on their
attitude or belief. In case of Online Airline Reservation Systems, it is suggested that
TAM and TPB should be considered with more belief-related variables. Therefore, in
our framework, self-determined psychological factors that may influence upon
perceived flexible personality of travelers by way of reinforcing their purchase
decisions are adapted from the TAM.
2.2.1 Usability Perception
Customers' acceptance of the internet, as a suitable medium for booking their
itineraries, has been accelerated due to the structural changes in the aviation industry
[97]. Likewise, research shows users dissatisfaction with existing Online Airline
Reservation Systems in terms of task completion (26]. Therefore, researchers are
eager to find out ways for improving online usability of the systems, how users'
Perceived Usability of the system is formulated by its flexibility functions [27]-[31].
In addition to techniques, methods and guidelines proposed for designing usable
systems, HCI researchers have also long argued on the importance of human factors
32
in designing and implementation of user-centered designs. According to Nielson [32],
"users experience usability of a site beji1re they have committed to using it and before
they have spent any money on potential purchases ... This indicates users' Perceived
Usability in online digital environments is an important determinant for evaluating
their satisfaction in the same environment. The existing literature can be divided into
four research aspects where researchers currently are focusing upon to determine
Perceived Usability of online systems.
2.2.1.1 Usability Perception by Performing Content Analysis
Many worldwide researchers in tourism industry have examined Perceived Usability
by performing content analysis of websites' features [98]-[1 03] which involves
technical assessment of the basic content and hypertext structure of the websites. For
example, Morrison eta/. [104] provided a comprehensive study of different websites
evaluation approaches by categorizing them into four groups. The grouping was based
upon determining "effectiveness and efficiency" and "Why and When" in evaluation
of the websites. Similarly, Law and Leung [101] examined 30 different North
American Online Airline Reservation Systems for evaluating the range of
comprehensive online reservation services provided by each of them. Their research
provides useful sets of attributes for consideration. Their findings which were based
on those useful sets of attributes showed significant differences user's satisfaction
with each website. Furthermore, Schar! eta/. [105] changed the traditional evaluation
techniques done by human experts and introduced an automated tool for the
systematically evaluation of websites.
2.2.1.2 Usability Perception through User's internet Adoption
The second area of research is based upon examining Perceived Usability of online
system through user's internet adoption practices [97]. One major pitfall of this kind
of research is that it primarily focuses upon use of internet technologies instead of
evaluating websites for determining Usability Perception of users. The investigation
33
methodologies typically involve interviews or surveys of tour operators, hotels and
destination marketing organizations [ 1 06], [ 1 07].
2. 2.1. 3 Usability Perception based on Users' Preferences and Expectations
The third area of research is based upon investigating web users' characteristics,
preferences, and expectations [1 08] from online systems, and to compare with the
Perceived Usability of the systems. For example, researchers have investigated the
demographic differences between the "lookers" and "bookers" [109], [110] in online
reservation systems and how their Perceived Usability different from one another due
to the contextual factors involved. Then there are a few academic studies where
researchers have investigated customer preferences and expectations for tourism
websites [97]. However, no study has been undertaken to examine customer
preferences and expectations from online airline reservation websites [97].
2.2.1. 4 Usability Perception based on Online Behavior of Web Users
The fourth most important and sophisticated area of research is the investigation of
online behavior of web users [97] and how their behavior is related to determining
perceived usefulness of the system. Some researchers have paid attention to online
users' search behavior on information [Ill], [112]. Bai eta/. [113] studied online
travel behavior of US college students. Rudstonn and Fagerberg [114] adopted quasi
experimental methodology to investigate customer's behavior and to explore
emerging concept of socially enhanced travel booking. In addition to this, Klein et al.
[115] and Marcussen [116] have examined online behavior of European travelers that
have possibly led to slower adoption trend of online air travel bookings in Europe.
Their findings showed that lack of relevant information, price dispersion, product
complexity, and the usability of online booking tools were the rationale behind
customers' such behavior.
34
2.2.2 Conceptual Linking between PU and FTB and FOARS
To discover the relationship between Perceived Usability (PU), Flexible Traveling
Behavior (FTB) and t1exible Online Airline Reservation Systems (FOARS), it is
necessary to study usability in the context of t1exibility from two different
perspectives. i.e. from user's traveling behavior and t1exibility of the reservation
systems.
Users traveling behavior is molded by a number of important personality relevant
determinants, both internal and external in characteristics. While, traveling
consciousness, self-efficacy in digital skills and self-belief as t1exible travelers are
internal personality relevant determinants int1uencing directly upon travelers t1exible
behavior, societal int1uences, attribution and prior experiences are external personality
relevant determinants that indirectly int1uence upon travelers t1exible behavior.
Moreover, external determinant may not necessarily always have the same int1uence
every time, depending upon the situation the traveler is in.
As for System's Flexibility is concerned, system's perceived t1exibility is
ret1ected in its Perceived Usability. Perceived Usability is a combination of system's
effectiveness, efficiency and satisfaction. End user support or user prompting is
considered to be a supporting characteristic that substantially augments efficiency and
effectiveness of the system, while multiple options directly int1uence upon user
satisfaction. If a system provides ranges of dates as t1ying and source destination
options at different fares, t1exibility of the system is enhanced in its Perceived
Usability in the eyes of the users. This is because if a user chooses a t1ying option 'A'
from a given one or two options, he has not made a t1cxible decision. But if he
chooses the same t1ying option 'A' from a variety of given t1ying options, he is likely
to enjoy extra satisfaction that he will get from the t1exibility of the system and also in
his purchase making decision. Likewise, if a system offers multiple t1ying options,
they will also int1uence upon users' decision and make them change their mind to opt
to t1y from option 'B' instead of 'A'. This will again have positive int1uence upon
user's satisfaction from the system- Perceived Usability. More details on this study
can be found in chapter 3 Section 3.3.1.
35
2.3 Usability V s. Flexibility
Researchers point out [117] that opaque products are flexible in characteristics;
therefore, a seller is in a unique position to offer horizontally differentiated products
to customers upon purchase due to the flexibility of assigning pre-determined
products to the customer. Opaque selling became popular since they offered a very
unique price discrimination mechanism [118] and could generate incremental revenue
for the airline by deliberating upon price sensitive consumers [119]. In very short
time, opaque selling has attained the status of a competitive lever for the airline,
signifying that an airline could suffer revenue loss to its competitors by not opting to
offer opaque offers [120].
The concept of online airlines reservation systems or SBTs for disposing off their
opaque inventory directly has not been adequately considered in information system
research and literature. As highlighted by Jerath eta/. [I 0] a number of studies have
carried out with primary focus upon airline n:venue management systems, however
attempts to empirically verify those findings are a few. However, some recent
research by Jiang [118], Fay [121], Granados e:t al [122], Jerath et al. [10] have made
strong argument in favour of this connotation and given theoretical justifications in
support of the argument. Likewise, recent r'~search on opaque selling has rather
adopted an objective discourse to empirically validate revenue management theories
(see, for example, Puller eta/. [123]). Granados eta/. [122], [124] has compared price
elasticities of the offline, online transparent and opaque channels. Their findings
suggest that opaque selling mechanism has high price elasticity. Again this does not
adequately address the research gap on opaque selling through SBTs.
It is believed offering opaque selling through SBTs will not reduce profits, as in
the case of direct last-minute selling undertaken by an airline, because it holds a
critical position to replicate its profit margin by setting high price of opaque selling.
But on a more fare note, airlines may like to introduce opaque selling directly through
their SBTs so as to attract additional sufficient number of customers and minimize the
effect of price degradation, as discussed earlier. Thus it is believed that if an airline
opts for opaque selling through its SBTs, it can enhance its profit through market
expansion and also by enhancing price discrimination of one's existing customer base
36
[ 121]. Moreover, in order to test applicability of opaque selling, empirical research on
opaque selling intermediaries show that it depends upon certain factors. e.g. demand
characteristics (Jerath eta/. [10], Fay and Xie [18]. Granados eta/. [122]), product
characteristics [I 18], consumer loyalty [ 121]. industry structure [120] and
competition [I 0], [ 125]. However. this research adopts an approach based on
behavioral characteristics. This is because literature review has highlighted that direct
last-minute selling, although could lead to severe consequences for the airline, is still
preferred over opaque selling through an opaque intermediary because of high
expectations of customers on little service differentiation [10]. And research on
opaque selling shows it has been preferred over direct last minute selling with an
increase in high demand situations and this has been the primary factor for airlines to
opt for opaque selling intermediaries [10]. However, if an airline opts for an opaque
fare selling it can position itself more competitively than opaque selling
intermediaries because of its knowledge and accurate resource information and
management [I 0]. Thus if an airline opts to adopt an opaque selling mechanism, it
will be in a win-win situation, whereby it will employ its own resources and provide
opaque selling directly to price-sensitive customers, who do not wish to anticipate
hidden characteristics of the service to be provided to them, a major concern in
opaque selling through intermediaries. Not only this, it will also add brand customer
loyalty to its credit as well.
The proposed model in this thesis thus integrates opaque selling mechanics into
the framework for designing of flexible Online Airline Reservation Systems. Thus, it
is important to discuss mechanics of opaque selling from same two perspectives
discussed earlier, System's Flexibility and User's Flexibility.
2.3.1 Opaque Selling Mechanics in System's Flexibility
In order to determine the role of opaque fare in making airlines SBTs (system)
flexible and improve their usability it is first and foremost important to discuss
flexibility of SBTs.
37
Earlier B2C systems were considered to be not very flexible as mentioned by
Olsen & Malizia [20], [21] in terms of handling more open requests that could not be
possibly mapped directly into the formalized terms offered by the Web-interface.
They would work for simple closed requests .. i.e., a request that could be mapped
directly into formalized terms or pre-defined parameters such as dates, airports,
flights, etc. The system could break down for more complex closed requests, i.e.,
where the customer is flexible with regards to attributes such as destination and dates
[20], [21]. Olsen & Malizia has recommended an information system, as an
intermediate between a customer and booking system, that would provide the user
with all the necessary data and support, on a mere button-click, after the initial data
has been fed into the system, see Figure 2.14.
_______ ... ·
Know-hew, experien:.e
--------- --- ---·
,--
I Book1ng System
Figure 2.14: Ticket Reservation through Intermediate
Source: Flexible User Interfaces for B2C Systems [20]
As per the above diagram, the role of intermediate system which is an SBT in this
case is to provide detail information to users, which could be useful in making good
decisions [20], (21]. Therefore, SBTs being intermediate information systems serve
the purpose of human agents in order to map formalized terms into closed requests.
And the existing SBTs cover mostly pre-sales flexibility [126]. The term "flexibility"
here should not only be related with the booking of a ticket in case of online booking
system. If we say, our booking/reservation system is flexible then ideally speaking, it
should also support flexibility with regards to flexible features of the system. Since
the essence of a successful quality website has been addressed by many authors from
time-to-time [74]-[78] and the quality of a website is referred to its usability and
functionality [24], [76]. Likewise, in case of usability of SBTs, users have reported to
be unsatisfied when they are flexible with regards to traveling attributes such as
destination and dates [20], [21], [26] and systems are not. That is why many
researchers are trying to find out ways so as to improve online usability [26]-[30] and
38
HCI researchers have long argued on the importance and criticality of human factors'
study to the successful design and implementation of technological devices so as to
improve system's usability by enhancing its flexibility.
In context of integrating opaque mechanics into existing SBTs of airlines, we
know from literature that when comparing last-minute direct sales of an airline, with
those offered through an opaque selling intermediary, researchers have found out that
opaque channel increases total demand [17] since customers are contended to
comprise on hidden attributes in anticipation of heavy discounts. This anticipation of
travelers for low fares is an extremely important concern, that airlines are faced with
every day. As mentioned earlier, airlines are not in a strategic position to offer low
cost fares directly to dispose-off their distressed inventories, mainly due to potential
threat to their revenue generation, therefore, the only business model that successfully
addresses this concern is opaque selling through intermediaries or OTAs. However by
opting for OT As, such as Price line for example, the traveling attributes such as the
airline or the route to be flown are hidden and research shows frustration experienced
by travelers when they end up flying a much more circuitous route, than they might
have wished and also not necessarily with the most preferred carrier. Moreover, even
if an airline sells its distressed inventory through opaque selling intermediaries, it
does not add brand loyalty of travelers to its credit, as travelers are likely to remain
impartial to the carrier they fly with under opaque selling [79]. Likewise, research
shows that more airline products become opaque or hidden in nature, higher the
dissatisfaction of the traveler becomes with the quality of airline solution or service
[79]. Therefore, it is argued that opaque selling mechanics if adopted by existing
SBTs, they will not only become flexible reservation systems and could even bring
similar or higher incremental revenue for the airline.
2.3.2 Opaque Selling Mechanics in User's Flexibility
Anticipation of travelers for low fares not only indicates why opaque selling
intermediaries became so popularly accepted, but also an idea about traveling
behavior of customers, which is 'the extent to which a traveler anticipates for a low
fare', indicating the extent to which 'a traveler is ready to compromise on flying
39
conditions', and thus becoming flexible in accepting what is being offered to them by
an airline. The idea of flexibility is not something new in airline industry, it was
proposed by Schwieterman [127] who had emphasized upon segmentation of the
market between discretionary and non-discretionary travelers using time flexibility.
More, recent research by Garrow [128], and Carroll et al. [129] also deploy time
flexibility as a value driver. However, there appears to be no research done to date, in
which destination flexibility is used as a value driver [79]. By designing SBTs in view
of User's Flexibility, they will be provided additional detailed information and
choices unquestionably useful in making good decision [20], [21].
2.4 Chapter Summary
As said earlier, this research is undertaken to examine applicability of opaque selling
mechanics on SBTs and have discussed the same in context of System's Flexibility
and User's Flexibility to increase the Perceived Usability of Online Airline
Reservation Systems. The first two research areas of Usability Perception i.e.
Usability Perception by Performing Content Analysis and Usability Perception through
User's Internet Adoption are related to supply oriented studies, the other two studies
i.e. Usability Perception based on Users' Preferences and Expectations and Usability
Perception based on Online Behavior of Web Users are demand driven since they
consider online services and features used by travelers when making traveling and
purchasing decisions. There are also some researchers who have combined some or
all four research areas for investigating Usability Perception of online systems. For
example, Benckendorff and Black [99] have used surveys of regional tourism
organizations as well as website evaluations methods. Nysveen [I 08] conducted
surveys and obtained results from both web users and tourism businesses. Their
research objective was to investigate gaps between customer preferences and actual
website offerings.
This research uses a blended approached and combines the four research areas of
Usability Perception to determine Perceiv(:d Usability of the Online Airline
Reservation Systems. It is important because customers' usability expectation and
preferences from Online Airline Reservation Systems lacks research and empirical
40
findings. Law and Leung [lOlJ had emphasized upon the need to investigate
expectations of airline customers that book their itineraries through their online self
booking tools. Moreover, the existing evaluation of online tourism websites is
performed by researchers and not by customers. lt leads to a dilemma and research
gap that does not potentially address expectations of travelers. Even though internet is
referred as a major technological innovation of today. its success heavily depends
upon assimilation of customer expectations and preferences into the design and
content of websites [1 01]. For airlines to run successful business, in spite of their
products are sold online or through more traditional channels, it is important to design
their self-booking tools in view of customers' preferences, expectations and online
usage behavior.
Thus, behavioral characteristics in terms of making travelers flexible are
appealing and an important area of research. And if it can be determined, i.e., what
makes a traveler flexible on the basis of his/her behavioral characteristics, it could
give crucial insight for designing of Flexible Online Airline Reservation Systems,
based on SBTs offering opaque selling. The existing opaque selling literature lies at
the intersection of consumer behavior and revenue management operational strategies
[ 40]. However no study has been found to address consumer behavior on opaque
selling with respect to online airline reservation system as most recent paper and
researchers are done within the marketing domain [ 40]. The research therefore
examines travelers' expectations, preferences and online behavior (User's Flexibility)
and aligns that with designing of flexible Online Airline Reservation Systems
(Systems Flexibility) and users' as evaluators of the online systems to determine its
usefulness through effectiveness, efficiency and satisfaction (Perceived Usability).
This anticipation of travelers for low fares is an extremely important concern, that
airlines are faced with every day. As mentioned earlier, airlines are not in a strategic
position to offer low cost fares directly to sell their left over inventories, due a number
of potential threats to their revenue generation, therefore, the only business model that
addresses this concern is opaque selling by OT As. The concept of online airlines
reservation systems or SBTs for disposing off their opaque inventory directly has not
been adequately considered in information system research and literature. The
concept requires extensive research especially in academic discipline [40] and as
41
highlighted by Jerath et a/. [ 1 OJ a number of studies have focused upon airline
revenue management systems, however attempts to empirically verify those findings
are a few. Likewise, recent research adopts an objective discourse to empirically
validate revenue management theories (see, for example, Puller et a/. [123]). This
does not adequately address the research gap on opaque selling. Moreover, the only
known studies on opaque selling are of Granados et a/. [122], [124] who have
compared price elasticities of the offline, onlline transparent and opaque channels.
Their findings suggest that opaque selling mechanism has high price elasticity.
Literature also highlights that direct last-minute selling, although could lead to
severe consequences for the airline, it is still preferred over selling through an opaque
intermediary in case of high expectations of customers and customer expecting little
service differentiation [10]. However, research on opaque selling also shows it has
been preferred over direct last minute selling with an increase in high demand
situations and this has been the primary factor for airlines to opt for opaque fare
intermediaries [10]. However, if an airline opts for an opaque fare selling it can
position itself more competitively than opaque selling intermediaries because of its
knowledge and precise information on availability of its own resources [I 0]. This
could put an airline in a win-win situation, whereby it will employ its own resources
and provide opaque selling directly to price-sensitive customers, who do not wish to
anticipate hidden characteristics of the service to be provided to them, a major
concern in opaque selling through intermediaries.
Anticipation of travelers for low fares, gives an idea about their traveling
behavior, which is 'the extent to which a traveler anticipates for a low fare',
indicating the extent to which 'a traveler is ready to compromise on flying
conditions', and thus becoming flexible in accepting what is being offered to them by
an airline. The existing opaque selling literature lies at the intersection of consumer
behavior and revenue management operational strategies [ 40]. However no study has
been found to address consumer behavior on opaque selling with respect to online
airline reservation system as most recent pap(:r and researchers are done within the
marketing domain [ 40]. This research therefore contributes to the small but growing
literature in designing of Online Airline Reservation Systems b y moulding upon
flexible behavior of travelers.
42
CHAPTER 3
METHODOLOGY
3.0 Chapter Overview
In this chapter, the overall methodology of the thesis is described which is divided
into three phases. Each phase contains one core research objective which is achieved
through the corresponding research questions and hypotheses. Section 3 .l is devoted
to the organization of the phases named as research methodology. Section 3.2 covers
Phase I that consists of 3 studies to investigate and discuss the user needs associated
to System's Flexibility and Users' Flexibility. Section 3.3 is dedicated to Phase II
which contains 2 further studies to classify users and to investigate the
interrelationship of the variables. At the end of Phase II. an overview of the study
results are also presented in order to facilitate the description of the methods used in
this research work. Section 3.4 presents Phase lii which contains a case study in order
to test the proposed framework. Section 3.5 presents the statistical formulas that are
used in this research and Section 3.6 summarizes the chapter.
3.1 Research Methodology
The methodology of this thesis consists of three phases which are described below.
The overall research methodology and a complete list of research questions and the
corresponding hypothesis are shown in Figure 3.1 and Table 3.1 respectively.
Phase 1:
Phase II:
Phase lll:
Assessing User Needs (System's Flexibility & Users' Flexibility)
User's Classifications (Interrelationship Testing of Variables)
Case Study (Testing the Model)
r-I I I I I
~r:~: An•ly,:-- . ----, r Analy5is ofE:\istmg Materials
Phase I
I r- ---- y -- j b -·- -----, I-S;udy3 -. -II ~~~tudy 1 Study.! , , :litv 1 • Flexibility in SBTs & 1 I l_~System s Flextbtllty _ • U~e-r~ Hex1b1~ : OTAs ___ _j I
L---------------------r------v--------~
U. sability Heuristic Evaluati~~l • Web-des1gn & Usab1lity
Guidelines • 3 Evaluators
' - --
I --- ___!'_'!' y ---- I Study 4 I
Usability of OARS L__ ________ _ I
I I I
~--- _ __,L_ ______ _
Flexibility Tacti(S Seledion • Literature • Bramstonning
3 HCI Experts Views in Questionnarire
Study 5 Users' Classilkation on the basis of Flexible Traveling Behavior
Phase II
________________ j
1---------------- -I
1 [; Interface B 1 (Integration of Opaque
I Mechanism into SBTs) I I I
Phase III
Validity Che~:~~x~riment)J-• Usability Testing
Figure 3.1: Research Methodology
44
~ v.
Phase Study
I
2
Research Objectives
To assess user needs (System's Flexibility and Users' Flexibility) associated with Online Airline Reservation Systems.
Table 3.1: Research Questions and Hypotheses
Research Questions Hypotheses
RQl: What are the issues H1: Non-functional Requirements are the with flexibility of Online perceived to have an impact on
Airline Reservation Systems, _u::::s:-:a.:..:bc.::i=li-'::tyc._.::_of::._cO-=.A.:::R=-=S.:.... -=--=-----whether or not flexibility is H 2: Functional Requirements are one of the reasons for users perceived to have an impact on the
not using such systems? ...:=u,:,:s.:::a.:.b.:,:il:"i t"'-y-'o=f.:...O=-:.cA=R-=-S=-=·--=-;-:--;;-;---::-c:::-c-=-:::-H3: The perceived flexibility of OARS affects the usability of such systems. H 4 : Functional Requirements of OARS are inversely associated with the flexibility of the systems. H 5: The availability of resources and skills set influence upon the usability of OARS.
RQ2: To what extend flexible H6: The level of satisfaction with users can compromise with existing SBTs is different for service quality attributes of respondents with different attitudes Online Airline Reservation towards Users' Flexibility in
Analysis
Correlation Analysis,
Reliability Analysis.
Descriptive Analysis.
Systems? compromising on SQAs of the airline. ANOVA & Post-H7: The level of satisfaction with Hoc existing OT As is different for respondents with different attitudes towards Users' Flexibility in compromising on SQAs of the airline.
~ 0\
Phase Study
3
II 4
5
Research Objectives
To propose a framework for designing more flexible Online Airline Reservation Systems while classifying users on the basis of their Flexible Traveling Behavior.
Table 3.1: Research Questions and Hypotheses (continues)
Research Questions Hypotheses
RQ3: How users' satisfaction H 8: Users' satisfaction with existmg with an existing SBTs is rated SBTs is different across their choice of against their choice of OT A four OT A features for making SBTs
Analysis
feature and reflected in their more FOARS. Two-Way ANOV A integration assessment of the same for making SBTs more flexible Online Airline Reservation Systems? RQ4: How users' perception on factors influencing flexible traveling behavior and flexible OARS is determined?
Emerging Theme Analysis
RQS: How to classify Users' H9: Users can be classified on the basis U .d. t' 1 . h . l 'bl f h . Fl 'bl T ]' B h . m Irec wna on the bas1s of t e1r F ex1 e o t e1r ex1 e rave mg e av10r. S 1 Traveling Behavior into High, ca e Medium and Low flexible and H 10: User's Flexible Behavior and their how to investigate Perceived Usability is correlated. interrelationships among
Correlation Analysis
System's Flexibility, Users' -H=1-1
:--,U;-;-s-er"""'·-s-;:F:;;-1-ex--,i::-b-;-l e---;:;B:-e-;-h_a_v:--i o_r_a_n--cd;-------C--l-.---Fl 'b']' d p . d I ·b·l· . l d orre atwn ex1 11ty an erce1ve System's F ex1 1 1ty 1s corre ate . A
1 .
U b.]. f . · 0 1· na VSlS sa 1 1ty o ex1stmg n me · Airline Reservation Systems? H 12 : Perceived Usability of OARS is not
affected by users' Flexible Traveling Behavior after adjusting for the effect of the covariate, System's Flexibility.
ANCOVA
~ __,
Phase
III
Study
Case Study
Table 3.1: Research Questions and Hypotheses (continues)
Research Objectives
To study the interrelationship between System's Flexibility, Users' Flexibility and Perceived Usability of Online Airline Reservation Systems and to determine the Perceived Usability of the existing and proposed systems.
Research Questions Hypotheses
RQ6: How do service quality Hn: Flexible behavior of travelers cannot attributes of airlines and be predicted by service quality attributes external variables jointly and external variables. predict flexible behavior of travelers?
RQ7: How does user H 14: User Perceived Usability with Perceived Usability with the existing and proposed systems is different existing and the proposed across the three groups. system differs?
RQ8: Is there a multivariate main effect of user's Flexible Traveling Behavior (High, Medium and Low) on effectiveness. efficiency and
Ht 5: There are differences among effectiveness, efficiency and satisfaction caused by the users' Flexible Traveling Behavior.
Analysis
Multiple Regression Analysis
Two-Way ANOVA
MAN OVA
~~--~~~------~~------~~--------satisfaction of the proposed H 16 : Effectiveness, efficiency and system? satisfaction in the proposed FOARS is
highest for users with highest flexible behavior.
Post-Hoc
3.1.1 Phase 1: Assessing User Needs (System's Flexibility & Users' Flexibility)
In Chapter 2, two different perspectives of flexibility, i.e. (1) System's Flexibility
(Computer) and (2) Users' Flexibility (Humans) were discussed. Therefore, in Phase
I, existing Online Airline Reservation Systems (Interface A) were used to assess the
System's Flexibility and Users' Flexibility. Three studies were conducted in this
phase as follow:
1. A study to investigate issues with flexibility and if flexibility is the reason for
not using Online Airline Reservation Systems.
2. A study to investigate users' flexible behavior in terms of compromising on
service quality attributes of an airline.
3. A study to examine if integration of OTAs features can make SBTs more
flexible in the context of Online Airline Reservation Systems?
Section 3.2 will discuss Phase I in detail.
3.1.2 Phase II: User's Classification (Interrelationship Testing of Variables)
Phase II was designed to carry out extensive relationship testing of variables and their
sub-measuring constructs towards developing a framework for designing a more
Flexible Online Airline Reservation System. Two detailed studies were conducted in
this phase with the following research objectives:
1. To explore the concept of users' perception on factors influencing Flexible
Traveling Behavior and Flexible Online Airline Reservation Systems.
2. To classify Users' on the basis of their Flexible Traveling Behavior (Low,
Medium, High) and to investigate interrelationships among System's
Flexibility, Users' Flexibility and Perceived Usability of existing Online
Airline Reservation Systems.
Section 3.3 will discuss Phase II in detail.
48
3.1.3 Phase III: Case Study (Testing the Proposed Framework)
Phase III is related to the design of a case study (paper prototype) and the
corresponding analysis to answer the final research questions. Participants were
requested to complete the usability evaluation of the existing and proposed interfaces
(Interface B). Quantitative technique was used to analyze the data collected in the
context of System's Flexibility, Users' Flexibility and Perceived Usability of the
systems. Section 3.4 will discuss Phase III in detail. The following research questions
were addressed:
l. How do service quality attributes of airlines and external variables jointly
predict the flexible behavior of travelers?
2. How does user Perceived Usability with existing and proposed systems
differs?
3. Is there a multivariate mam effect of user's Flexible Traveling Behavior
(High. Medium and Low) on effectiveness, efficiency and satisfaction of the
proposed system?
3.2 Phase I: System's Flexibility and Users' Flexibility
Three studies were conducted in this phase as shown below using the existing OARS
to assess the System's Flexibility and Users' Flexibility.
3.2.1 Study 1: Issues with Flexibility
This study is to address the I st research question.
RQI: What are the issues with flexibility of Online Airline Reservation Systems,
whether or not flexibility is one of the reasons for users not using such systems?
49
3.2.1.1 Rationale
The rationale behind pilot study I was to assess users' needs from the System's
Flexibility perspective. The primary objective of Requirements Engineering (RE) is to
build a foundation of a product that satisfies the customers' needs and interests. In
addition to pure functionality, customers want to see non-functional characteristics in
the systems, such as security, stability, usability and high performance. Such Non
Functional Requirements (NFRs) have become essential for the success [130] of
today's businesses.
The role of NFRs for the success or failure of any Transaction Processing System
is as important as in any other systems. In the case of electronic commerce, security
and privacy of consumers' sensitive personal data are one of the major concerns
[131]. Consumers' lack of full adoption of electronic commerce solutions is not
merely due to the concern on security and privacy, but also caused by a variety of
non-functional characteristics such as flexibility, consistency, learnability, and
reliability. This study specifically addresses the notion of flexibility of such systems,
trying to establish how much the success of Business to Consumer (B2C) e-commerce
hinges on them.
The main objective of pilot study I in phase I was to investigate customers'
concerns for not using Online Airline Reservation Systems, and to study the
relationship between flexibility and the use of Online Airline Reservation Systems.
Therefore, this study focuses on the following sub research questions:
• What are the reasons for not using Online Airline Reservation Systems?
• Do customers have concerns with the flexibility of current Online Airline
Reservation Systems?
• Does the perceived flexibility of Online Airline Reservation Systems
significantly affect the usability of such systems?
50
3. 2.1. 2 Methodology
In this phase, the methodology of the study followed quantitative research based on
the self-reporting questionnaire and testing of hypothesis as shown in Figure 3.2. The
hypotheses were:
H 1: Non-functional Requirements (NFRs) arc perceived to have an impact on the
usability of OARS.
Hz: Functional Requirements (FRs) are perceived to have an impact on the
usability of OARS.
H3: The perceived flexibility of OARS affects the usability of such systems.
H4: Functional Requirements of OARS are inversely associated with the
flexibility of the systems.
H5: The availability of resources and skills set influence upon the usability of
OARS.
r::z__-,J I FloKll]ty j Col:li.qen;;y I L«rrl.d:>!IJiy [lM.M:lity I Seruri:v 1
EasytJ use I
_i~ (:1_____ __ ~ Ill '
~
Rt-qum·d Ro:sotUTtS ,~
.'lkill:i-1
I H~ +
i ~-----
~-\(, ,. u~bililr oe Cltln~ '"" \_ Airllll(> RP~I'V:'Iholl J
SY";ft>m<; ~ //
/ -- ....... ~,----~
~t
Figure 3.2: Research Model for the Hypothesis to be Tested
51
3.2.1.3 Validity
A questionnaire was designed in order to test the model and the above hypotheses
(Appendix A). A mixture of two approaches-adaptive (questions from existing
literature) and development (questions as per the required scenario) - was used to
prepare the questionnaire. Relevant and useful questions were adapted from the
literature review [132]-[134].
In order to ensure that items on the questionnaire were related to the constructs
being measured, the Content Validity Index (CVI) was used. According to Bums and
Grove [135] Content Validity is obtained from three sources: literature,
representatives of the relevant populations, and experts. The Content Validity Index
(CVI) developed by Waltz and Bausell [136] was used in this study.
Three human factor students with expertise in usability were asked to rate each
item on the questionnaire based on Relevance, Clarity, Simplicity and Ambiguity on
the four-point scale. The results of CVI were analyzed and items that had CVI over
0.75 remained and the rest were discarded. The: remaining items were modified, based
on the experts' opinions.
The questionnaire consists of 13 questions on Non-functional Requirements -
NFR (excluding flexibility), 9 questions on Functional Requirements - FR, 11
questions exclusively on Flexibility, and 3 questions on assessing Users' resources &
skills when booking or using Online Airline Re:servation Systems.
The pilot testing of the questionnaire was also conducted through a series of
informal interviews with PhD and Master level students at Universiti Teknologi
PETRONAS before sending the questionnaire to the target audience.
3.2.1.4 Sample Size
To analyze the model under consideration and to study the factors that impact the use
of online reservation systems, more than 200 copies of the questionnaire were emailed
or hand-distributed to various students and faculty members of two Universities
52
namely, Universiti Teknologi PETRONAS (UTP) and Universiti Teknologi Mara
(UiTM). The analysis of the participants is shown in Figure 3.3.
r!>er's.Anal_ys.i> (In:stiJ:ution)
Srudy 1
"C"ni;-·~1~in
T::okr,oli::·~i
P;:t:ror.::.!, -:2.4~~
Figure 3.3: User's Analysis of Participants from UTP
3. 2.1. 5 Response Rate
140 responses were received, yielding a response rate of 70% of the total population
surveyed. Out of 140, 78 respondents reported to be inexperienced with online ticket
buying; the remaining 62 respondents considered themselves experienced.
3.2.1.6 Scale
Each construct included questions presented in a five-point Likert mode, ranging from
"strongly agree" to "strongly disagree." Respondents' responses were scored as: for
the "strongly agree" response was assigned a score of 5, while for the "strongly
disagree" response was assigned a score of I. Consequently, users' gaining higher
scores in a certain scale showed stronger preferences toward the specific scale.
3.2.1. 7 Analysis
In order to test influence of variables on another, 5 research hypotheses were designed
in this study that was correlated using Pearson Correlation Coefficients. Pearson
Correlation (Section 3.5.2) is regarded as the most familiar measure of examining
dependence between two quantities. It indicates the strength of a linear relationship
53
between two variables; however its value generally does not completely characterize
their relationship [ 13 7]. Additional analyses were performed on data, by computing
descriptive analysis (Section 3.5.1) and Reliability Analysis (Section 3.5.3). Data
analysis carried out will be discussed in Chapte:r 4.
3.2.2 Study 2: Users' Flexible Behavior in Terms of Compromising on SQAs
This study is to address the 2"d research question.
RQ2: How flexible users are in term of compromising on Service Quality Attributes
(SQAs) of Online Airline Reservation Systems?
3. 2. 2.1 Rationale
The rationale behind study 2 was to assess users' needs from the Users' Flexibility
perspective. A mixture of two approaches ~ adaptive (questions from existing
literature) and development (questions as per the required scenario) ~ was used to
prepare the questionnaire (Appendix B). Relevant and useful questions were adapted
from the literature review [134), [138), [139].
3.2.2.2 Methodology
Study 2 followed a quantitative research methodology based on self-reporting
questionnaire and testing of hypothesis. The null and alternate hypotheses for the
study are given below:
H6: The level of satisfaction with existing SBTs is different for respondents with
different attitudes towards Users' Flexibility in compromising on service quality
attributes of the airline.
54
H0: The level of satisfaction with existing SBTs is same for respondents with
different attitudes towards Users' Flexibility in compromising on service quality
attributes of the airline.
H7: The level of satisfaction with existing OT As is different for respondents with
different attitudes towards Users' Flexibility in compromising on service quality
attributes of the airline.
H0: The level of satisfaction with existing OT As is same for respondents with
different attitudes towards Users' Flexibility in compromising on service quality
attributes of the airline.
3.2.2.3 Validity
For the second pilot study, after ensuring Content Validity using Waltz and Bausell
[136] scale, Criterion Validity was computed through Concurrent validity to ensure
whether the questionnaire is truly measuring users' satisfaction with existing SBTs
and OTAs.
In psychometrics, Criterion Validity is a measure of how well one variable or set
of variables predicts an outcome based on information from other variables, and will
be achieved if a set of measures from a personality test relate to a behavioral criterion
on which psychologists agree [140]. Criterion Validity was ensured through
implementing concurrent validity of the measuring constructs, i.e. Users' Flexibility
in terms of compromising on service quality attributes of airlines (Users' Flexibility)
and effectiveness, efficiency & satisfaction (Perceived Usability), by taking feedback
of 12 randomly selected users. Concurrent validity is particularly useful to
demonstrate where a test correlated with a measure has previously been validated
[141]. In this case, Pilot study I had already established a strong correlation of Users'
Flexibility with Perceived Usability of Online Airline Reservation Systems (Hs).
Validity check results showed a strong positive correlation of r = 0.354, p < 0.05
between Users' Flexibility in terms of compromising on service quality attributes of
55
airlines and Satisfaction. Essentially, this means Users' Flexibility in terms of
compromising on service quality attributes of airlines can be used to predict their
satisfaction from existing online reservation systems, both SBTs and OTAs.
3.2.2.4 Sample Size
250 copies of the questionnaire were hand-distributed to various students at Universiti
Teknologi PETRONAS and four travel agencies in Malaysia located in the city of
Ipoh, Perak. The analysis of the participants from travel agencies and those
respondents from Universiti Teknologi PETRONAS are shown in Figure 3.4 and
Figure 3.5 respectively.
User's Analysis- Travel Agencies (TA) Study 2
RA Jits Travel & Tours, Adil Travel
12% ~ & Tours, :jjlg;;,, 33%
Kinta Valley \j ~mlii_M('".· Travel •f0~~1~f:
Agency, · \ffiL'
32% . .a...:: Taikar
~-Holidays, 23%
Figure 3.4: User's Analysis of Participants from Travel Agencies
User's Anal<>ysis (UTP) Study2
Figure 3.5: User's Analysis of Participants from UTP for Study 2
56
3. 2. 2. 5 Response Rate
169 responses were received, yielding a response rate of 67.60% of the total
population surveyed. Out of 169, 106 (63%) respondents were from Universiti
Teknologi PETRONAS, while the remaining 62 ( 3 7%) respondents were travelers
visiting traveling agencies.
3.2.2.6 Scale
Each construct included questions presented in a five-point Likert mode, ranging from
"strongly agree" to "strongly disagree." Respondents' responses were scored as: for
the "strongly agree" response was assigned a score of 5, while for the "strongly
disagree" response was assigned a score of l. Consequently, users' gaining higher
scores in a certain scale showed stronger preferences toward the specific scale.
3.2.2.7 Analysis
In order to test the stated alternate and null hypothesis H" and H1 in this study, One
way Analysis of Variance (ANOVA) was computed to determine the satisfaction
mean of users with existing SBTs and OT As. ANOV A (see Section 3.5.4) is a
statistical method used to compare the means of two or more groups.
ANOV A for H6 was computed to determine satisfaction mean of users with
existing self-booking tools of the airline, who at the same time reported their
flexibility level in terms of compromising on service quality attributes of the airline.
The respondents had to select from the three given options of, (I) Can compromise on
service quality attributes, (2) May compromise on service quality attributes, (3)
Cannot compromise on service quality attributes. This was followed by ANOV A for
H7 to determine satisfaction mean of users with existing online travel agencies of the
airline, who at the same time reported their flexibility level in terms of compromising
on service quality attributes of the airline. The respondents had to select from the
three given options of, ( l) Can compromise on service quality attributes, (2) May
57
compromise on service quality attributes, (3) Cannot compromise on service quality
attributes. Data analysis carried out will be discussed in Chapter 4.
3.2.3 Study 3: Integration of OT As Features can make SBTs more FOARSs
This study is to address the 3'd research question.
RQ3: How users' satisfaction with an existing SBTs is rated against their choice of
OT A feature and reflected in their integration assessment of the same for making
SBTs more flexible Online Airline Reservation Systems?
3.2.3.1 Rationale
The rationale behind study 3 was to investigate if the integration of some OT A
features could make SBTs a more Flexible Online Airline Reservation Systems and
how that can be achieved? Therefore, a preliminary but comprehensive focus group
research was conducted with airline executives, using quantitative survey method to:
• examine their subjective satisfaction with existing self-booking tools provided
by the airlines,
• report if the group approve or disapprove the proposed idea of the four OT A
features (Matrix Display, Opaque Fare, Alternate Airport Search, Hotel Search
Facility) integrated into SBTs so as to make them Flexible Online Airline
Reservation Systems, and
• recommend an OTA feature of their choice, for making SBTs Flexible Online
Airline Reservation Systems.
3.2.3.2 Methodology
Study 3 followed a quantitative research methodology based on self-reporting
questionnaire to test the hypothesis. The study used the same data set as obtained in
58
study 2. Therefore, the steps required for validity, sample size, response rate and scale
used in the methodology can be seen in study 2.
The null and alternate hypothesis for the study is given below:
Hs: Users' satisfaction with existing SBTs is different across their choice of four
OTA features for making SBTs more Flexible Online Airline Reservation Systems.
H0: Users' satisfaction with existing SBTs is same across their choice of four
OTA features for making SBTs more Flexible Online Airline Reservation Systems.
3.2.3.3 Analysis
In order to test if the respondents approve or disapprove the idea of OT A features
integrated into SBTs so as to make them Flexible Online Airline Reservation System,
means plot was examined. This was followed by two-way ANOV A analysis to test
the satisfaction level of existing Online Airlines Reservation Systems. One of the
basic assumptions before performing any analysis of variance is to check for
normality of sampling distribution of mean. The sample size for this study (n) was
169, and according to central limit theorem if a random sample of size n is> 30 and it
is derived from an infinite population with finite standard deviation, then the
standardized sample mean converges to a standard normal distribution [142].
To perform a Two-way ANOVA, respondents were requested to indicate if
integration of OT As features will make Online Airlines Reservation System more
flexible or not and also to recommend an OTAs feature (from the given four options
of Opaque Fare, Matrix Display, Hotel Search Facility, Alternate Airport Search) for
integration into SBTs.
The F -statistic was interpreted in analysis of variance since it is a ratio of the
explained variability to the unexplained variability (taking into account the degrees of
freedom). A larger F-statistic indicates that more of the total variability is accounted
for by the model [143]. Data analysis carried out will be discussed in Chapter 4.
59
3.3 Phase II: Users' Classification and Intenelationship Testing of Variables
Two studies were conducted in this phase to classify users on the basis of their
flexible traveling behavior.
3.3.1 Study 4: Users' Perception on Factors Influencing Flexible Traveling
Behavior
This study is to address the 41h research question.
RQ4: How users' perception on factors influencing Flexible Traveling Behavior and
Flexible Online Airline Reservation Systems is determined?
3.3.1.1 Rationale
The rationale behind this study was to investigate and explore the concept of users'
perception on factors influencing Flexible Traveling Behavior and Flexible Online
Airline Reservation Systems.
3.3.1.2 Methodology
In order to gather an in-depth understanding of human behavior and the reasons that
govern such behavior, qualitative research was conducted. This qualitative
exploratory study adopted a grounded theory approach to investigate the users'
perception on their Flexible Traveling Behavior and Flexible Online Airline
Reservation Systems. Grounded Theory is a research method in which the theory is
developed from the data, rather than the other way around [ 144], since it is an
appropriate way to research a previously little studied area in Information Systems
research. According to Strauss [145], "A grounded theory is one that is inductively
derived from the study of the phenomenon it represents. That is, it is discovered,
developed, and provisionally verified through systematic data collection and analysis
of data pertaining to that phenomenon. "
60
Moreover. this methodology provides an ideal and flexible guideline to analyze
qualitative data and equip researchers with necessary understanding of underlying
concepts to build theories through successive levels of data analysis [ 146].
Researchers [ 14 7]-[ 149] have recognized this method as an authentic research tool in
qualitative data analysis due to its procedural credibility.
The population of this study consisted of travelers who had expenence m
purchasing tickets through airlines Self-Booking Tools (SBTs) and Online Traveling
Agencies (OTAs). This was an important consideration, because travelers with
experience in purchasing tickets through SBTs and OT As could very well understand
and relate to what being 'Flexible Traveling Behavior' mean from users' perspective
as well as from systems' perspective.
The data was collected from three methods. (I) Two online travel forums,
http://www.travelblog.com and www.travellerspoint.com/ (2) Semi-structured in
depth interviews and (3) Focus group. Use of online surveys to collect data has
become a popular choice for researchers, with special reference to tourism data [150).
This was mainly due to the flexibility, reach and robustness offered by visual medium
of internet. Likewise, in depth interviews and focus group were essentially required in
this research, so as to build deeper understanding of respondents' perspective on
Flexible Traveling Behavior and Flexible Online Reservation Systems, which
otherwise may not be possible to obtain through online travel forums alone.
Moreover, in short interviews researcher is in a position to pick up non verbal cues
and even rephrase questions so as to personalize them and make respondents feel at
ease to answer them.
The following questions were raised at travel forums:
• Which factors influence upon your Flexible Traveling Behavior?
• Which factors influence upon your perception of a Flexible Online Airline
Reservation System?
61
The semi structured in-depth interview of actual travelers was conducted at Kuala
Lumpur International Airport from 11-l3th March, 20 II. A realistic, flexible and
ethically accepted approach was adopted to identify potential research participants.
Since this research involved travelers and reservation systems, therefore, Kuala
Lumpur International Airport (KLIA) was visited for consecutive three days in order
to approach a number of volunteer travelers.
Finally, the focus group interviews were conducted with 3 managers, 5 junior
executives and 3 technical experts of three local airlines, namely (I) Malaysian
Airline, (2) Fire Fly, and (3) Air Asia. The interviews were held from 15-25th March,
2011.
3.3.1.3 Validity
In order to ensure study's trustworthiness.. two methods were employed I.e.
Triangulation and Negative Case Analysis.
With regards to triangulation, the three sources and three different data collection
methods; online travel forums, semi-structured in-depth interviews and focus group,
were employed. This was important to see that data obtained from different
independent data sources converged on something similar, or at least do not oppose to
each other [151]. The data was analyzed by two authors independently and then
discussed together to derive emerging themes, categories and to also ensure
credibility. Negative case analysis was perfmmed on the initial derived emerging
themes [146], [152]. The purpose was to see: if the characteristics of the derived
emerging theme sufficiently inculcated the tme essence of whole research and were
applicable to all cases.
3.3.1.4 Sample Size
31 respondents of the two questions were from travelblog's and travellerspoint
forums. 28 travelers were interviewed, and each interview lasted from I 0-15 minutes.
62
Finally, the focus group interviews were conducted with 3 managers, 5 junior
executives and 3 technical experts of three local airlines, namely (I) Malaysian
Airline, (2) Fire Fly, and (3) Air Asia. The interviews were held from 15-25th March,
20 II (permission letter enclosed see Appendix F). The analysis of the participants
from the two online travel forums, semi -structured in-depth interviews and the Focus
group are shown in Figure 3.6, Figure 3.7 and Figure 3.8 respectively.
User's Analysis (Online Travel Forums) Study ..t
point 55%
Travelblog 45%
Figure 3.6: User's Analysis of Participants from Online Travel Forums
User's Analysis (Semi-Structured in-depth Interviews)
S d Study 4 M .. u an auntms
Australia----- 4% -------- 4% 4% --- \ Thailand ___ _
7% Malaysia 21%
7%
UK :__-:::,..,..--- I I%
China 7%
Figure 3.7: User's Analysis of Participants from In-depth Interviews
User's Analysis (Focus Group) Study 4
Aero Asia-. ~Malays. ian 27% Airline ~ 37%
FireFly~ 36%
Figure 3.8: User's Analysis of Participants from Focus Group
63
3.3.1.5 Response Rate
43 responses were received from online travel forums. Out of 43, 12 cases were
rejected due to the ambiguous respondents yielding a response rate of 72% of the total
responses. 28 responses in-depth interviews and II responses from focus group were
collected with no rejected cases. The demographics of online travel forums, in-depth
interviews and focus group can be seen in Chapter 4.
3.3.1.6 Scale
The analysis of interview transcripts was based on an inductive approach which is
meant to identify emerging patterns in the data by using thematic codes. Inductive
analysis looks for emerging patterns, themes and categories through analysis of data
and opposes imposition of the same, prior to data collection and analysis [153].
3.3.1. 7 Analysis
The data collected from three different sources was examined for triangulation. It
depicted a similarity pattern, especially in case of data collected from online travel
forums and in-depth interviews of travelers at KLIA.
Data analysis of later, however, provided a more detailed perspective of travelers
flexible behavior by incorporating socio-economic factors and societal influences.
The focus group, being technical experts, however significantly contributed towards
identifying factors that may influence upon perceived flexibility of reservation
systems. After giving much thought process to results as shown in Chapter 4, Section
4.4.1, 6 themes emerged under factors influencing upon Flexible Traveling Behavior
and 3 themes emerged under factors influencing upon perceived flexibility of a
reservation system as shown in Table 3 .2.
64
Table 3.2: Emerging Themes on Factors lnl1uencing upon FTB and PF
Emer in Themes
Factors Influencing Upon Flexible Traveling Behavior
Factors Influencing Upon Perceived Flexibility (PF) of an
Online Airline Reservation System 1. Travelers' flexible behavior is moulded 1. Systems perceived flexibility is
by their traveling consciousness. 2. Travelers' J1exible behavior is moulded
by their belief that they have the required digital skills.
3. Travelers' l1exible behavior is moulded by their self-belief as flexible travelers.
4. Travelers' J1exible behavior is moulded by societal influences.
5. Travelers' flexible behavior is moulded by how they attribute a cause to their traveling behavior.
6. Travelers' flexible behavior is moulded by their prior traveling experiences.
inl1uenced by its Perceived Usability.
2. Systems perceived flexibility ts infl uenccd by end user support.
3. Systems perceived J1exibility ts inl1uenccd by companson of features on the actual level of effect regarding to complete the reservation process.
3.3.2 Study 5: A Study to ClassifY Users' on the Basis of Flexible Traveling
Behavior
This study is to address the 51h research question.
RQS: How to classify Users' on the basis of their Flexible Traveling Behavior into
High, Medium and Low J1exible and how to investigate interrelationships among
System's Flexibility, Users' Flexibility and Perceived Usability of existing Online
Airline Reservation Systems?
3.3.2.1 Rationale
The study purpose was to (i) classify users' on the basis of their Flexible Traveling
Behavior (Highly Flexibile, Medium Flexible, Low Flexible) and (ii) to investigate
interrelationships among System's Flexibility, User's Flexibility and Perceived
Usability of the Online Airline Reservation Systems.
65
3.3.2.2 Methodology
The quantitative research methodology was adopted usmg survey to address the
following four hypotheses.
H9: Users can be classified on the basis of their Flexible Traveling Behavior.
H10: User's Flexible Traveling Behavior and their Perceived Usability IS
correlated.
H 11 : User's Flexible Traveling Behavior and System's Flexibility is correlated.
H12: Perceived Usability of Online Airline Reservation Systems is not affected by
users' Flexible Traveling Behavior after adjusting for the effect of the covariate,
System's Flexibility.
3. 3. 2. 3 Validity
In order to ensure that items on the questionnaire were related to the constructs being
measured, the Content Validity Index (CVI) was used. According to Burns and Grove
[135] Content Validity is obtained from three sources: literature, representatives of the
relevmt populations, and experts. The Content Validity Index ( CVI) developed by
Waltz md Bausell [136] was used in this study.
Three humm factor students with expertise in usability were asked to rate each
item on the questionnaire based on Relevmce, Clarity, Simplicity and Ambiguity on
the four-point scale. The items that had CVI over 0.75 remained md the rest were
discarded. The remaining items were modified, based on the experts' opinions.
3.3.2.4 Sample Size
To investigate the above hypotheses, 90 random cases were selected for validating the
results of trmsformation scale and to perform preliminary interrelationship of
variables before performing the final analysis on the data set of 273 responses.
66
3.3. 2. 5 Response Rate
Out of 273, 90 random cases were selected as shown m Figure 3.9. Randomly
selected cases will be used for transformation scale.
L
Travelers 183
67%
User's Analysis (Transformation Scale) - -1 Study 5
Random Selection for Transfonning
Scale 90
33%
Figure 3.9: User's Analysis of Participants for Transformation Scale
3.3.2.6 Scale
Five-point Likert mode was used rangmg from "strongly agree" to "strongly
disagree." Respondents' responses were scored as: for the "strongly agree" response
was assigned a score of 5, while for the "strongly disagree" response was assigned a
score of 1. Consequently, users ' gaining higher scores in a certain scale showed
stronger preferences toward the specific scale.
3. 3. 2. 7 Analysis
Section B of the questionnaire consisted of questions on Service Quality Attributes
(SQAs) of an airline that users' are ready to forgo or compromise on, in order to
become flexible travelers. By opting to become flexible travelers, users would get
discounted fares at the cost of being unaware of three hidden characteristics of their
traveling itinerary (1) their seat details/confirmation, (2) date of flying confirmation
and (3) time of flying confirmation. All questions posted in this section were uni
directional and designed as such that the extent to which a user would agree with a
67
particular statement reflected his/her flexible behavior in terms of flying on flexible
dates/times. The following unidirectional scale was used, where I denoted 'Highest'
and 5 denoted 'Least' flexibility in terms of compromising on service quality
attributes of an airline as shown in Table 3.3.
Table 3.3: Uni-directional Scale to Measure Users' Flexibility
Neutral Low Least Rate your pnorities for the following Highest High service quality attributes of an airline in 1-----.----.-----.----.-----1 terms of their importance. I 2 3 4 5
I Flying date confirmation 2 Flying carrier confirmation 3 Flying time confirmation 4 Number of stop-over 5 Number of connected flights 6 Ticket class (economy/business) 7 Seat specifications 8 Discounted Airfares 9 Destination/Source Airport
Immediate confirmation of 10
itinerary on purchase of ticket
Transformation Scale - Once the score of the respondents was recorded, it
required adoption of a validated mechanism to test the hypothesis H9, whereby
individual ratings of participants on I 0 service quality attributes could be transformed
into singular unit, such as a score or product.
In a study conducted on aesthetic appraisal of 5 most popular aviation websites in
the world, a similar methodology based on transformation of ratings into a product
score, has been used [154]. Respondents in the study evaluated 5 popular SBTs of
airlines on the given parameters by assigning a score of I (Very Low) to 5 (Very
High). Then their score was transformed into a product, by examining frequency of
occurring of each rating (I, 2, 3, 4, 5) and multiplying that frequency (e.g., I occurs
two times, 2 occurs zero times, 3 occurs four times, 4 occurs four times, 5 occurs zero
times) with that of transforming scale product. The final total is added to obtain a
unique product score, which in case of given example 2 as shown in Table 3.4. Data
analysis carried out will be discussed in Chapter 4.
68
Table 3.4: Transformation Scoring Scale
-----
Rating VL L N H VH I 2 3 4 5 -- --
Scale -2 -1 0 +I +2 Rating Frequency 2 0 4 6 0 Product (Rating Frequency* Scale) -4 0 0 6 0 Product Score 2
Hto investigated interrelationship between users' Flexible Traveling Behavior
(FTB) and their Perceived Usability of FOARS. For this investigation, Kendall's Tau
(see Section 3.5_9) method for examining bivariate correlations was selected because
it essentially met the nonparametric conditions of the study. First, the study used a
small data set of 62 respondents only, reporting their Perceived Usability. Secondly,
users were classified into three categories of flexible behavior (high, medium, low) on
the basis of their Users' Flexibility score. The correlation analysis was performed and
significance level was interpreted. Data analysis carried out will be discussed in
Chapter 4.
H 11 investigated interrelationship between users' FTB and System's Flexibility
(comprised of adaptability, adaptivity, and personalisation). Pearson Correlations was
computed and significance level was checked and interpreted accordingly. Data
analysis carried out will be discussed in Chapter 4.
H 11 called for the further investigation in order to ascertain interrelationship
between users' Flexible Traveling Behavior and System's Flexibility. Consequently,
in H 12, Perceived Usability of Flexible Online Airline Reservation Systems (PU) was
included as a dependent variable to investigate how the two variables, ( l) Users'
Flexible Traveling Behavior and (2) System's Flexibility together predict Perceived
Usability of Online Airline Reservation System. Moreover. since the interrelationship
between users' Flexible Traveling Behavior and Perceived Usability of Flexible
Online Airline Reservation Systems is already ascertained in H 10, therefore, users'
Flexible Traveling Behavior was taken as a fixed factor and System's Flexibility was
included as a Covariate, to reduce within group error variance and to eliminate
confounds.
69
ANOV A includes one or two continuous variables that predict the outcome or
dependent variable. However, continuous variables such the one that are not part of
the main experimental manipulation but may have an influence on dependent variable
are known as co variates in the Analysis of Covariance (AN COY A). ANCOV A takes
into account confounding variables to give a clear measure of effect of the
experimental manipulation, and the analysis is performed as such to first examine
influence of independent or fixed factor (users' FTB) on dependent variable
(Perceived Usability of Flexible Online Airline Reservation Systems) and then
experiment is manipulated by introducing a covariate (System's Flexibility).
One important assumption was checked before performing ANCOV A, 1.e.,
independence of the covariate and treatment effect. This assumption requires that
covariate (System's Flexibility) should not be different across three users' Flexible
Traveling Behavior Groups in the analysis. To meet this assumption, One Way
Independent ANOV A was performed, with Perceived Usability of Flexible Online
Airline Reservation System across three groups as an independent variable, and
System's Flexibility as an outcome variable. This analysis should be non-significant
to meet the assumption of ANCOV A. This was followed by performing ANCOV A
results are interpretation. Data analysis carried out will be discussed in Chapter 4.
3.3.3 Research Way Forward
The Matrix display and hotel search features are unique in the sense of incorporating
multiple sources reservation information, however from integrating into the proposed
framework of SBTs perspective, they are not very feasible since it requires merger of
multiple information resources, that might not be acceptable to an airline due to its
privacy policy and other regulations. On the contrary, alternate airport search is
related to provide additional information, and the extent to which a traveler is willing
to be flexible in identifying his/her destination sources. This feature seems practical
and has implications for integration into SBTs. Finally, unlike other OTA innovations,
the opaque fare mechanism depends on hidden characteristics of the traveling plan,
thus leveraging upon traveling behavior of leisure travelers, who are always up for
70
grabs and less sensitive to traveling plans. The findings of the studies showed that this
OT A feature to be the most recommended for integration into SBTs as shown in
Figure 3.1 0.
I Atrlm~ ~cs-cr:~~tlon I Sv~~=~n: __ _
y_ ____ _
I Extstmg SBT
__ _!ea~~-
~---,.Opaque Fares
Matrix Display
Alternate Atrport Search
Hotel Search
y ___ _
Proposed SATs J Figure 3.10: Proposed FOARS after Integrating Opaque Fares into SBTs
Since we are investigating flexibility from both, system point of view and users'
perspective, integration of opaque fares concept into SBTs would increase the
usability of the system by way of improving System's Flexibility and by making
users' more flexible in their decision making as shown in Figure 3.11.
SBTs L___ __ - - _____ _!
: Adaptabllity --- ---~
.,· Adaptlvtty I!
Personaltzat•on
···--- --, ' DeCISIOn Makmg I
4 , Aehavior
Figure 3.11: Increased Usability by Improving SF and Users Flexible Decision
As discussed above, flexibility is referred to its ability to respond to internal or
external changes. Change can be defined as the transition over time which requires
change agent. Researchers argued that "if the change agent is external to the system,
then the change under consideration is a flexible-type change" [12]. Therefore, in case
of SBTs incorporated with opaque fares would serve the role of external change agent
71
by way of providing flexibility in users' decision making. Similarly, "if the change
agent is internal to the system, then the change under consideration is an adaptable
type change" [12). Thus the provision of opaque fares into SBTs also serves the role
of internal change agent by way of providing the capability of accepting changed
decisions. If no change agent exists, then the system is rigid (no change can occur).
Since provision of opaque fares could make users flexible and also increases the
adaptability of the system, it is expected that the usability of the system would be
enhanced.
3.4 Phase III: Case Study to Test the Propos1:d Framework
Phase III was designed to study the interrelationship between System's Flexibility,
Users' Flexibility and Perceived Usability of Online Airline Reservation Systems and
to determine the Perceived Usability of the existing and proposed systems. In order to
investigate the relationship between Perceived Usability and travelers Flexible
Traveling Behavior (FTB) with existing and proposed systems the following three
research questions were investigated:
• RQ6: How do Service Quality Attributes of airlines and External Variables
jointly predict flexible behavior of travelers?
• RQ7: How does user Perceived Usability with existing and proposed system
differs?
• RQ8: Is there a multivariate main effect of user's Flexible Traveling Behavior
(High, Medium and Low) on effectiveness, efficiency and satisfaction of the
proposed system?
3.4.1 Rationale
The case study was designed in order to investigate the relationship between user
Perceived Usability and travelers Flexible Traveling Behavior with the existing and
the proposed Online Airline Reservation Systems.
72
3.4.2 Methodology
The quantitative research methodology. based on investigating three (03) research
questions through survey questionnaires was adopted. Users' were given hard copies
of the questionnaire with following sections.
Section A of the questionnaire was designated to collect demographic profile of
the respondents. Section B consisted of service quality attributes of an airline that
users' are ready to forgo or compromise in order to become flexible travelers. By
opting to become flexible travelers, users will get discounted fares at the cost of not
knowing their seat confirmation, date of flying confirmation and time of flying
confirmation. All questions posted in this section were uni-directional and designed as
such that the extent to which a user would agree with a particular statement reflected
his/her flexible behavior in terms of flying on flexible dates/times. Section C
consisted of questions that particularly addressed the external factors that may
influence upon the Perceived Usability of the system. The details of each measuring
construct have already been discussed in Chapter 2. Section D consisted of questions
to measure Perceived Usability of the system.
To answer the above research questions the following hypotheses were created:
• Hn: Flexible behavior of travelers cannot be predicted by the Service Quality
Attributes and External Variables.
• H 14: User Perceived Usability with existing and proposed systems is different
across the three groups.
• H 15: There are differences among effectiveness. efficiency and satisfaction
caused by the users' Flexible Traveling Behavior.
• H 16: Effectiveness, efficiency and satisfaction m the proposed FOARS IS
higher for users with highest flexible behavior.
73
3.4.3 Validity
Validity technique known as Item Discrimination Index was employed in this study
[155]. Item Discrimination Index indicates how adequately an item separates or
discriminates between high scorers and low scorers on an entire test [157]. It is a
measure of difference between the proportion of high scorers answering an item
correctly and the proportion of low scorers answering the item correctly. Andy Field
[ 156] argues that item discrimination means that respondents with different score
should also differ in the construct of researchers' interest. Kelley [158] suggested that
item discrimination should be based upon following two corollaries and pose
unidirectional questions to respondents, so that the degree of their agreement with a
particular statement could be used to discriminate them with respondents with certain
levels of disagreements over the same statemem.
• Respondents with the same score should be equal to each other along the
measured construct.
• Respondents with different scores should be different to each other along the
measured construct.
To meet the item discrimination validity requirement, the questionnaire was
designed with uni-directional questions. This paved way to discriminate respondents
on the basis of the degree to which they agree being flexible travelers or not.
3.4.4 Sample Size
To investigate the above three (03) research questions, more than 500 travelers were
requested to fill in the survey questionnaire during the Malaysian Association of Tour
and Travel Agents (MATTA) fair 2011. MATTA fair is a world known exhibition in
the tourism industry which is held every year. Three (03) days (ll-13'h March 2011)
were spent at Putra World Trade Centre (PWTC), Kuala Lumpur (Appendix G).
74
3.4.5 Response Rate
273 responses were received, yielding a response rate of 54.6% of the total population
surveyed. However, 23 cases were dropped due to the ambiguous respondents
yielding a response rate of 50%.
3.4.6 Paper Prototype
A paper prototype as shown in Figure 3.12 and Figure 3.13 was developed to get the
users' response to classify them into different categories.
P.lon Tue Wed Thu Fro Sat Sun .--2 -• 3- 4 _( S I
6 J. 7 J""i--, 9 I 10 II ~~ 13 _I 1~ I IS 16 \ 17 18 1 19 ~ 21 _! 22 23 I :z~ :s .20 27 -128l'2FF'30 _____ _
Figure 3.12: Prototype of a Flexible Booking Window for Flexible Travelers
• •
Figure 3.13: Prototype of a Flexible Booking Window for Inflexible Travelers
75
3.4.7 Usability Evaluation of Prototype
The usability of a prototype named as interface B was inspected through heuristic
evaluation. 3 HCI graduates who had usability evaluation and testing experiences
served as expert evaluators to evaluate the usability. An evaluation sheet was
provided to each evaluator that contained a series of usability guidelines based on the
well-recognized Research-Based Web Design and Usability Guidelines book [159].
The selected guidelines were adapted to be presented in a checklist format (Appendix
H). The evaluators were requested to assign each guideline a problem severity rating
from 0 (no problem) to 4 (usability catastropht~) based on the frequency, impact and
persistence of the problem. According to Nielsen and Mack [ !60), using the mean of a
set of severity ratings from three evaluators is satisfactory for many practical usability
inspection purposes.
Furthermore, the evaluators were also requested to provide redesign suggestions
for each problem identified. Based on all evaluators' evaluation and the results from
studies, a new interface B (Prototype) was designed.
3.4.8 Scale
Questionnaire was designed using psychometric scales, which are commonly used in
psychological research [79). This technique was employed because psychometric
scales tend to prompt an individual to respond to various questions that pertain to a
given context and according to Davis [79) "responses of individuals are an indication
of their internal belief'". The participants were provided a hardcopy of the
questionnaire for indicating their response by evaluating the prototype by using a 5
point Iikert scale ranging from (I) "Strongly Disagree" to (5) "Strongly Agree"
(Appendix C).
3.4.9 Analysis
The analysis on RQ6, RQ7 and RQ8 are as follows:
76
Analysis of RQ6- The first research question in the case study was investigated
by performing statistical procedure in two steps. In Step I. Pearson Correlation
Coefficients of the ten quality service attributes along with external variables was
computed to determine their association with flexible behavior of travelers and to also
ascertain their individual range and strength of association. In Step 2. Multiple
Regression Analysis (MRA) was performed to determine how service attributes
quality and external variables jointly determine flexible behavior of travelers. MRA
predicts values on a quantitative outcome variable, using several other predicting
variables [161). Data analysis carried out will be discussed in Chapter 4.
Analysis of RQ7- In order to investigate the second research question of the case
study i.e. how Perceived Usability of existing and proposed systems differs among
users' with high, medium and low Flexible Traveling Behavior?, a two-way ANOVA
analysis was performed. As mentioned above, one of the basic assumptions before
performing any analysis of variance is to check for normality of sampling distribution
of mean. The sample size for this study (n) was 250, and according to central limit
theorem if a random sample of size n is > 30 and it is derived from an infinite
population with finite standard deviation, then the standardized sample mean
converges to a standard normal distribution [142].
Two-way ANOV A tested the Effectiveness. Efficiency and Satisfaction of the
existing and proposed systems for travelers with High, Medium and Low Flexible
Traveling Behavior. The F -statistic was interpreted in analysis of variance since it is a
ratio of the explained variability to the unexplained variability (taking into account the
degrees of freedom). A larger F-statistic indicates that more of the total variability is
accounted for by the model [143].
Analysis of RQ8 - Third research question of the case study was examined with
the help of MANOVA to see if there is a multivariate main effect of user's Flexible
Traveling Behavior on the proposed system's effectiveness, efficiency and
satisfaction. Multivariate normality requires that any linear combination of the
dependent variables must be distributed normally. This assumption was checked by
examining pair wise nonlinear relationships between dependent variables using scatter
77
plots. The second assumption of multivariate analysis is homogeneity of the
covariance matrices. It was met by examining Box's M, which tests the hypothesis
that the covariance matrices of the dependent variables are significantly different
across levels of the independent variable as show in Chapter 4, Section 4.8.1.2.
The overall F test for the three dependent variables was examined in Multivariate
Tests by analyzing the statistic called Wilks' lambda (;\,), and the F value associated
with that which is significant at p <00 1. Lambda is a measure of the percent of
variance in the Dependent Variables that is *not explained* by differences in the level
of the Independent Variable. If the overall F test is significant, then it is a common
practice to go ahead and look at the individual dependent variables with separate
ANOV A tests. This was followed by a univariate ANOV A and Post-hoc multiple
comparison tests that shows statistically significant effect on three dependent
variables. Data analysis carried out will be discussed in Chapter 4.
3.5 Statistical Methods
In this thesis, different statistical methods have been chosen to analyse the data. The
methods were, descriptive analysis, pearson coefficient correlation, reliability
analysis, analysis of variance, analysis of covariance, multivariable analysis of
variance, multiple regression analysis, Post Hoc Scheffe's Test and Kendall's Tau.
The following subsection describes the statistical formulas used in this research.
3.5.1 Descriptive Analysis
Descriptive analysis is used to describe and summarize a collection of data in a clear,
understandable and meaningful manner which allows simple interpretation of the
data. There are two basic approaches, one is numerical and, second is graphical.
Using the first approach i.e. numerical method, one might compute statistics such as
mean and standard deviation. In graphical method one might create plots that contain
detailed information about the distribution. Stern and leaf display and a box plot are
famous plots in graphical method [162).
78
3.5.2 Pearson Coefficient Correlation
The correlation coefficient was developed by Karl Pearson from the idea introduced
by Francis Galton in 1880s [163], [164]. Sometime, it is also termed as "Pearson's r".
In statistics, correlation (linear dependence) measures the degree of association
between two variables X and Y. It is a value between -I and +I inclusive. A positive
value implies a positive association between the two variables (large values of X tend
to be associated with large values of Y and small values of X tend to be associated
with small values of Y). A negative value implies a negative or inverse association
between the variables (large values of X tend to be associated with small values of Y
and vice versa). It is defined as the covariance of the two variables divided by the
product of their standard deviations.
3.5.3 Reliability Analysis
In statistics, reliability is treated as the consistency of a set of measurements or it is
also referred as a measuring instrument which is used to describe a test. It is inversely
related to the random error (165]. Reliability is said to be sample dependent as it is the
property of the scores of a measure rather than the measure itself. Cronbach's alpha is
the most common consistency measure which is typically interpreted as the mean of
all possible split-half coefficients (166]. Cronbach's alpha is a more generalized form
of Kuder-Richardson Formula 20 [166] used for estimating internal consistency.
Reliability may be defined as the proportion of true score variability that is captured
across respondents, relative to the total observed variability [167].
3.5.4 ANOVA
The aim of Analysis of Variance generally known as ANOVA is to test the significant
difference between group means. ANOV A is commonly used if the user needs to
compare performance of more than two parameters. ANOVA generalizes !-test to
more than two groups by providing a statistical test of whether or not the means of
several groups are all equal. Therefore. the advantage of ANOV A over t-test is
79
ANOV A can detect the effect of interaction between variables, and to test more
complex hypothesis about the existing problem [168]. If the result indicates a
significant difference, then it would be followed by post-hoc test to identify which
mean of result is different.
ANOV A is a technique that was firstly proposed by R. A. Fisher in a 1918 article
"The Correlation Between Relatives on the Supposition of Mendelian Inheritance". In
1921, his first application of the ANOV A was published which was later included in
his book "Statistical Methods for Research Workers". ANOV A is used to compare
group means [169]. ANOVA uses two hypotheses to determine the result, namely null
hypothesis and alternate hypothesis.
3.5.5 ANCOV A
The aim of Analysis of Covariance generally known as ANCOV A is to compare one
variable in 2 or more groups taking into account variability of other variables, called
covariate [170]. ANCOVA is a technique that sits in between analysis of variance and
regression analysis [ 171]. It combines one-way or two-way ANOVA with linear
regression. In other words, it is a General Linear Model (GLM) with a continuous
outcome variable and two or more forecaster variables where at least one is
continuous and one is categorical. Continuous variable is always quantitative or
scaled while the categorical variable is nominal or non-scaled.
3.5.6 MANOV A
Multivariate analysis of variance generally known as MANOV A is a generalized form
of ANOV A which is used to analyze data that involves two or more than two
dependent variables [172]-[175]. There are three basic advantages of MANOVA
analysis. Firstly, it helps in finding the inte:ractions among dependent variables,
secondly, it helps in finding the interaction among independent variables and, thirdly,
it helps in finding the effect of independent variable(s) on the dependent variable(s)
[176].
80
3.5.7 Multiple Regressions
The term regression was first time introduced by Francis Galton in 1900s [ 177).
Galton's work was later extended by lldny Yule [178] and Karl Pearson [179] to a
more general statistical context. In statistics [161J. this technique is used to predict the
relationship between one dependent variable and one or more independent variable(s).
This technique is able to form a method of least square where sum of squared
residuals between the regression plane and the observed values of the dependent
variable are minimized [180].
3.5.8 Post Hoc Scheffe's Test
Post hoc tests are useful to explore the differences among means. It provides specific
information on which means are significantly different from each other. Therefore,
Post hoc tests are performed where researchers has already conducted F -test with a
factor that consists of more than two means [181]. In statistics, there are many
procedures to perform Post hoc tests, however, Scheffe's techniques is a most popular
and flexible method introduced by Henry Scheffe. It is a method to adjust significance
levels in linear regression analysis for multiple comparisons.
3.5.9 Kendall's Tau
Kendall's Tau is used to measure the strength of the relationship between the two
variables. It is a measure of correlation which is carried out on the ranks of the data
[182]. For any sample ofn observations, there are [n (n-1)/2] possible comparisons of
points (XI. Y1) and (XJ. Y1). Assume C is a number of pairs that are concordant, and D
is a number of pairs that are not concordant than Kendall's Tau can be calculated
[183).
81
3.6 Chapter Summary
In this chapter, we discussed the overall methodology of the thesis whereby existing
Online Airline Reservation Systems were examined to assess the flexibility and
usability of the systems. This chapter was divided into three phases and each phase
contains one core research objective which was achieved through quantitative and
qualitative techniques to assess System's Flexibility, Users' Flexibility and Perceived
Usability of the systems. A redesign solution for enhanced usability for more Flexible
Online Airline Reservation Systems was developed based on HCI guidelines and the
flexibility tactics used in online travel agencies. A new Flexible Online Airline
Reservation System design was applied, which led to a proposed interface with the
integration of opaque mechanism. The two interfaces were used in the case study.
Participants were requested to complete the evaluation of the existing and proposed
interfaces.
82
CHAPTER4
RESULTS AND ANALYSIS
4.0 Chapter Overview
The research methodology discussed in Chapter 3 was divided into three phases
addressing one core research objective with corresponding research questions in each
phase. The same pattern is followed in this chapter to organize the results obtained
through the corresponding hypotheses. To achieve the I st research objective, Section
4.1, 4.2 and 4.3 are devoted to presents the results obtained from the three studies
conducted in Phase I answering the corresponding research questions RQ I, RQ2 and
RQ3, respectively. To attain the 2"ct research objective, Section 4.4 and 4.5 are
dedicated for the results obtained from the two studies conducted in Phase II
answering the corresponding research questions RQ4 and RQ5, respectively. To
conquer the 3'ct research objective, Section 4.6, 4.7 and 4.8 presents the results
obtained from the case study answering the corresponding research questions RQ6,
RQ7 and RQ8, respectively. Section 4.9 summarizes the overall chapter.
4.1 Phase 1: User Needs Associated to System's Flexibility
This study is to address the I st research question.
RQl: What are the issues with flexibility of Online Airline Reservation Systems,
whether or not flexibility is one of the reasons for users not using such systems?
4.1.1 Descriptive and Reliability Analysis
72% of the online experienced ticket buyers found the systems to be consistent with
respect to the usage of terms and the position of messages on the screen. Among the
experienced buyers, 71% agreed that the current online reservation systems facilitate
learning through textual descriptions and are presented in a manner that is not
confusing. 35% of the experienced users found online reservation systems were easy
to use or handle with little need to read instructions. 61% of the participants agreed to
have no difficulty in learning to operate the system.
Pearson correlation coefficients were computed in order to test the relationships
between each factor and the usability of online reservation systems. Table 4.1 shows
the average item scores and standard deviations within each of the four groups (FR,
NFR, flexibility and required resources/skill set).
Table 4.1: Coefficient of the Factors Affecting Usability of Online Systems
Factors No. of Mean per
S.D. Cronbach's a Items Factor
Functional Reguirements 9 1.6 0.7 0.74 Non-Functional Reguirements 20 2.8 1.0 0.76 Consistency 5 2.2 0.9 0.90 Ease of Use 5 3.2 0.9 0.62 Learnability 4 3.8 1.1 0.93 Security 3 3.4 0.8 0.63 Trust 3 3.6 0.9 0.71 Flexibility 11 2.8 0.7 0.81 Satisfaction 4 3.2 0.8 0.74 User-guidance 4 3.2 0.8 0.76 Simplicity 3 2.3 1.0 0.82 Required Resources & Skill Set 5 3.1 0.6 0.66
The largest group-wise score (3.1 with a standard deviation of 0.6) was attained
for Required Resources and Skills. Even though this is almost the mean of the five-
point Likert scale, this indicates that this factor was of high preference among
customers. Ranked next were Flexibility (2.8 with a standard deviation of 0.7) and the
other NFRs (2.8 with a standard deviation of 1.0), respectively. This close ranking
suggests that a notable preference of customers concerning these two aspects. The
lowest item score is that of the FRs (1.6 with a standard deviation of 0.7), indicating
that the extent to which online reservation systems may provide incorporation of
additional support features such as online cancellation, online modification, online
84
transfer and online reservation, remams uncertain as per the perception of the
customers.
Cronbach's a analysis was performed to assess the reliability of the variables used
in the research constructs. Reliability analysis was computed on each section/group
measuring different attributes associated with the hypothesized research constructs.
• Flexibility: This factor gained the highest Cronbach's a score of 0.81
• NFR: This factor gained the second highest Cronbach's a score of0.76
• FR: This factor gained the third highest Cronbach's a score of0.74
• Required Resources and Skills: This factor gained Cronbach's a score of0.66
4.1.2 Hypothesis Testing Hr
H 1: Non-functional Requirements are perceived to have an impact on the usability of
OARS.
The factor of NFR was differentiated into separate dimensions to capture the
customer's perception of consistency, ease of use, learnability, security, and trust.
Table 4.1 shows that the construct "Leamability" among the NFRs had the highest
mean of3.8 and a standard deviation of 1.1, which may indicate an inclination of the
users toward learning how to handle advanced features of online reservation systems.
Moreover, this construct also has the highest Cronbach's a score of 0.93. The items in
the Learnability construct were related to the satisfaction of users with reading text on
the screen, the sequence of screens, the organization of information and supportive
information such as online help, messages, and documentation provided by the
systems before, during, and after completing the tasks. This is followed in rank by the
extended construct of "Trust" with a mean of 3.6 and a standard deviation of 0.9. The
extended construct of Trust was related to the risk in providing personal accounts
information online and the reliability on experienced travel agents in finding better
flights and packages. The survey indicates that customers bestowed upon traveling
agents greater trust and reliability believing they could help in finding them better
85
traveling packages. This construct has the third highest Cronbach's a score of 0.71.
The extended construct of "Consistency" has the second highest Cronbach's a score
of 0.90. The questions asked with regards to this construct included various aspects of
how information is presented in the interface and used by customers of online
reservation systems. The items in the Consistency construct addressed finding the
consistency in use of terms throughout the system, the positioning of messages on the
screens, organization of screens, and the use of text-based instructions.
The correlation between NFR and the usage of the systems was observed to be r =
0.7 with p<O.Ol. This shows the significance of the relationship; the hypothesis is
accepted.
4.1.3 Hypothesis Testing H2
H2: Functional Requirements are perceived to have an impact on the usability of
OARS.
Table 4.2 presents the flexibility of existing systems. 66% of the experienced
users stated that they never tried to make online changes in their traveling schedule.
15% of the respondents reported to have been successful, II% reported to have been
unsuccessful, and the remaining 8% reported to have not seen such an option in
existing online systems. 64% of the respondents claimed that they never tried online
cancellation. 8% reported to have been successful in making cancellation changes,
16% reported to have been unsuccessful, whereas the remaining 12% reported to have
not seen such an option in the systems. 80% of the respondents claimed that they
never tried an online transfer of a ticket. I% reported to have been successful, II%
reported to have been unsuccessful, whereas the remaining 8% reported to have not
seen such an option in the system. 72% of the respondents answered that they never
tried online correction of errors. 9% reported to have been successful, 15% reported to
have been unsuccessful, whereas the remaining 4% reported to have not seen such an
option in the system. Lastly, 58% of the experienced users reported that they never
tried reservations in online systems. Interestingly, no one reported to be successful,
whereas the remaining 15% reported to have not seen such an option in the system.
86
Table 4.2: Flexibility of Existing Systems
Flexibility of Existing Systems Never Unable Option not
Successful to do so available
Have you ever tried to make changes in your traveling dates 66% II% 8% 15% online? Have you ever tried to cancel
64% 16% 12% 8% your ticket online? Have you ever tried to transfer your ticket to someone else 80% II% 8% 1% online? Have you ever tried to correct
72% 15% 4% 9% typos errors online? Have you ever tried to reserve a ticket for few days with the 58% 27% 15% 0% intension to buy it later?
The relationship between FR and the usage of the systems was observed to have r
= 0.18 with p>O.OI. This shows that the relationship is not significant; the hypothesis
is rejected.
4.1.4 Hypothesis Testing HJ
H3: The perceived flexibility of OARS affects the usability of such systems.
The factor of flexibility was investigated from different viewpoints to examine the
perception of the customers concerning satisfaction, user guidance, and simplicity of
use. 45% agreed that the data entry was flexible, 35% perceived flexible user
guidance, and 53% saw clear indications for completing the process. 40% claimed to
be satisfied with the total number of steps required to accomplish the task.
Furthermore, 48% were satisfied with the ease of completing the tasks and 41% with
the total time systems take to complete the tasks. 30% of the respondents agreed that
the online systems support quick and easy recovery from mistakes, while 32% of the
respondents did not agree. The remaining 38% of the respondents remained neutral to
this. 30% of the respondents agreed that the current online reservation systems
provide effective linkages with other travel-related partners (e.g. links to other airline
reservation system in case of connected flights). 25% did not agree, whereas the
remaining 45% remained neutral on this. 67% of the respondents claimed to have
87
been uncertain about how to fix errors if these occur; 85% of the respondents
preferred travel agents to make changes in their flight schedule. 72% of the
respondents preferred travel agents for reserving tickets. 51% of the inexperienced
respondents refrained from online shopping so as to avoid making online payments.
51% of the respondents considered travel agents to be more reliable. Two extended
constructs of the "Flexibility" factor, namely, "User-guidance" and "Satisfaction"
have the highest mean score of 3.2. The reliability of both the constructs is also
observed to be good with a Cronbach's a score of0.76 and 0.74, respectively.
The relationship between perceived flexibility of online reservation systems and
the usage of such systems was observed to be r = 0.69 with p<O.O 1. This shows a high
significance of the relationship; the hypothesis is accepted.
4.1.5 Hypothesis Testing H4
H4: Functional Requirements of OARS are inversely associated with the flexibility of
the systems.
The relationship between FR of online reservation systems and the flexibility of
the systems was observed to have r = 0.28 with p<O.Ol. This shows that the
relationship is mildly significant; the hypothesis is accepted.
4.1.6 Hypothesis Testing Hs
H5: The availability of resources and skills set influence upon the usability of OARS.
10% of the respondents claimed to have no Internet connection available; 30%
claimed to have no or little knowledge of online reservation systems; the remaining
60% reported to have no credit or debit card available.
The relationship between available resources and the usage of online reservation
systems is observed to haver= 0.32 with p<O.Ol. This shows that the relationship is
mildly significant; the hypothesis is accepted.
88
4.2 Phase 1: Users' Flexibility in Terms of Compromising on SQAs
This study is to address the 2"d research question.
RQ2: To what extend flexible users can compromise with service quality attributes of
Online Airline Reservation Systems?
4.2.1 Assumptions in ANOV A to Test H6
Before performing ANOV A, a very basic assumption of ANOV A was checked, i.e.
absence of outliers. Box-plot of the sample distribution was examined since it is a
useful standard in data interpretation, reveals data symmetry, skewness and the
presence of outliers. Moreover, it also facilitates in comparing more than one
population without knowing anything about the underlying statistical distributions of
those populations.
4.2.1.1 Box-and-Whisker Plot (H6)
In case of satisfaction level with existing SBTs, respondents who reported that they
'can compromise' on SQAs of the airline have a median at 3 (black line) as shown in
Figure 4.1.
Satisfaction Level With existmg Self Bookmg Tools (SBTs)
I - H 1ghly Sa11sfiod, 5- H 1ghly Dtssallsfied
6
5 5 .... Cll eo
" " a 4
-5 3: 3
]
j 2
i Cll 1
0 N • ~ n ~
C1n Comprom1se May Com pmmiso Cannot Com prom1se
Users' fleJUbllll y 1n compromISing o n sernce q u.aluy annbul es
Figure 4.1: Box Plot showing Satisfaction Level with Existing SBTs
89
This represents neutral satisfaction level and at the same time indicates 50% of the
data is greater than this value. Users' with any lesser satisfaction with existing SBTs
are represented everything above median black line, while the users with higher
satisfaction are represented everything below median black line. As shown by the top
'whisker', this group has greatest values but no outliers. Hence the data is normally
distributed.
In case of satisfaction level with existing SBTs, respondents who reported that
they 'may compromise' on SQAs of the airline have a median at 2 (black line). This
represents high satisfaction level and at the same time indicates 50% of the data is
greater than this value. Users' with any lesser satisfaction with existing SBTs are
represented everything above median black line, while the users with higher
satisfaction are represented everything below median black line. As shown by the top
'whisker', this group has greatest values and an outlier. The majority of the data is
normally distributed.
In case of satisfaction level with existing SBTs, respondents who reported that
they 'cannot compromise' on SQAs of the airline have a median at 2 (black line). This
represents high satisfaction level and at the same time indicates 50% of the data is
greater than this value. Users' with any lesser satisfaction with existing SBTs are
represented everything above median black line, while the users with higher
satisfaction are represented everything below median black line. As shown by the top
'whisker', this group has greatest values but no outliers. Hence the data is normally
distributed.
4.2.1.2 Means Plot (H6)
The means plot as shown in Figure 4.2 shows that there is apparently an enormous
difference between the satisfaction level of the three respondents groups, which
appears not be the actual case. Therefore as a follow-up, the same results will be
analyzed in a different chart to see the difference between the groups.
90
Satisfaction leve 1 V.'ith existing Self Booking Tools (SBTs)
I - lilghly Satisfrcd, ~ Highly Drssattstied
32r-------------------------------~
§ 2.4
~ § :1] 2.2
2.0 .1----------------------------------:-1 Can Compromrse May Compromrse Cannot Compromise
Users' flexibrlrty m compromismg on servrce quality attributes
Figure 4.2: Means Plot on Satisfaction Level with Existing SBTs
4.2.1.3 T-Test (H6)
In this case the three groups are significantly different usmg a t-test (t=36.760,
df=169, p=O.OOO) as shown in Table 4.3. 95% Confidence Interval (CI) is probability
that the interval contains the true mean.
Table 4.3: T-Test on User's Flexibility with SBTs
TestValue=O ----· -··-· -------
t df Sig. (2- Mean 95% Confidence Interval tailed) Difference of the Difference
------.--
How flexible 36.760 169 .000 2.07
Lower Upper you are 1.96 2.18
4.2.1.4 Error Bars (H6)
The same results are now reproduced in the error bars, with 95% confidence intervals
to have an idea of the variation in sample distribution. CJ of the groups is closely
related to the results of the analysis of variance for these groups. The confidence
interval for each graph below shows a linear pattern of the sample distribution which
otherwise appeared to be showing huge variations in the simple means plot.
91
In error bars we intend to see if the mean of one group is included in the
confidence interval of the other two groups - if so then there is likely no difference
among the groups. Moreover, it is not relevant whether the error bars 'overlap' but
whether the mean of one group 'overlaps' with the error bars of the other. The
confidence intervals can overlap by as much as 25 percent of their total length and
still show a significant difference between the means for each group.
Error Bar: Satisfaction level with existing Self Booking Tools (SBTs)
I- Highly Satisfied, 5- Highly Dissatisfied
~ 3.5
"' "' ~ 3.0
1 ~ 2.5
" .Q s • 2.0 • "' I I I
1.5
'0 " ~·a " Can Compromise Cannot Compromise
May Compromise
Users' flexibility in compromising on service quality attributes
Figure 4.3: Error Bar on Satisfaction Level with Existing SBTs
In Figure 4.3, 95% CI tells us that the satisfaction level of existing SBTs for the
users who "can compromise" on SQAs of the airline is probably between 2.7 and
3.35, with group mean of3. Likewise, for users who "may compromise" it is probably
between 2.4 and 2.83, with group mean of 2.6, and for users who "cannot
compromise" it is probably between 1.8 and 2.3, with group mean of2.07.
The group means of users' who 'may compromise' shares a degree of confidence
interval overlap with users who 'can compromise', thus the two groups may not
necessarily be different from one another. Moreover, the group mean of users' who
'cannot compromise' does not share any degree of confidence interval overlap with
either of the two groups, therefore, this particular group appears to be significantly
different from the rest of the sample population. However, post-hoc tests can confirm
this.
92
4.2.2 Hypothesis Testing H6
H 6: The level of satisfaction with existing SBTs is different for respondents with
different attitudes towards Users' Flexibility in compromising on SQAs of the airline.
4.2.2.1 One-way Analysis of Variance (lh)
To test the hypothesis, one-way analysis of variance was used to determine the
satisfaction mean of users with the existing self-booking tools of the airline and at the
same time report their flexibility level in terms of compromising on SQAs of the
airline as shown in Table 4.4.
Table 4.4: One-Way ANOV A on Satisfaction Level with Existing SBTs
Sum of Squares df Mean Square F Sig. Between Groups 20.707 2 I 0.354 11.250 .000 Within Groups 153.699 167 .920 Total 174.406 169
The respondents had to select from given three options of, (I) Can compromise on
SQAs, (2) May compromise on SQAs, (3) Cannot compromise on SQAs. The
analysis showed significant differences among satisfaction mean of the three user
groups with existing self booking tools of the airline (F (2,169) = 11.250, p < .001).
The respondents who indicated their flexible attitude as "cannot compromise" on
SQAs of the airline, depicted highest level of satisfaction with existing SBTs of the
airline (M = 2.06, SD = .916). This was closely followed by satisfaction of the
respondents who indicated their flexible attitude as "may compromise" on SQAs of
the airline (M= 2.59, SD = .973). The respondents who reported their flexible attitude
as "can compromise" on SQAs of the airline, depicted least level of satisfaction with
existing SBTs of the airline (M = 3.00, SD = .987).
Since the three user groups differed significantly on satisfaction mean level of
existing SBTs of the airline, null-hypothesis is rejected and alternative hypothesis H6
is accepted.
93
4.2.2.2 Post-hoc Scheffe Tests(H6)
Post-hoc Scheffe tests in Table 4.5 showed that there is a significant difference
between the pair of means of the respondents who reported their flexible attitude as
"Cannot compromise" on SQAs of the airline with those who "Can compromise";
p=.OOO(<.OOI). The same group of respondents also differed significantly from the
group of respondents who reported their flexible attitude as "May compromise" on
SQAs of the airline, p=.009 (<.01).
Table 4.5: Multiple Comparisons on Satisfaction Level with Existing SBTs
Dependent Variable: Satisfaction with Existing SBTs
(I) How (J) How Flexible
Mean Std.
95% Flexible You
You Are Difference
Error Sig. Confidence
Are {1-J) Interval Lower Upper Bound Bound
Can May Compromise .41 .187 .092 -.05 .87
Compromise Cannot
.94(*) .202 .000 .44 1.44 Comrromise
May Can Compromise -.41 .187 .092 -.87 .05
Compromise Cannot
.53(*) .172 .009 .II .96 Com2romise
Cannot Can Compromise -.94(*) .202 .000 -1.44 -.44
Compromise May Compromise -.53(*) .172 .009 -.96 -.II
* The mean difference is significant at the .05 level.
4.2.2.3 Effect Size for One-way ANOVA(H6)
From hypothesis testing it is clear that the three groups are different, but this does not
confer the strength or the magnitude of this effect. Effect size is measure of the
strength of an effect. And since the null-hypothesis has already been rejected,
therefore it makes sense to calculate effect-size to determine the size of the effect. The
size of the effect is 12% (IJ2 = 0.1187).
94
4.2.3 Assumptions in ANOV A to Test H7
Box-plot of the sample distribution was examined to meet the assumption of
ANOV A, i.e. absence of outliers. Since it is a useful standard in data interpretation,
reveals data symmetry, skewness and the presence of outliers. Moreover, it also
facilitates in comparing more than one population without knowing anything about
the underlying statistical distributions of those populations.
4.2.3.1 Box-and-Whisker Plot (H7)
In the case of satisfaction level with existing OT As, respondents who reported that
they 'can compromise' on SQAs of the airline have a median at 2 (black line) as
shown in Figure 4.4. This represents high satisfaction level and at the same time
indicates 50% of the data is greater than this value. Users' with any lesser satisfaction
with existing OT As are represented everything above median black line, while the
users with higher satisfaction are represented everything below median black line. As
shown by the top 'whisker' , this group has greatest values but no outliers. Hence the
data is normally distributed.
40
0~ 3 5 ~ .,. ·B a,a .:; i 25
] 2.0 g ·c ~ 15
lii
Sati<;faction Level with existing Online Travel Agencies (OT As)
I • Higly Satisfied, S -lh!#lly DISsatisfied
"' 1.0
N11: <2 83 fJI5
Can Comprorruse May Compromise Cannot Compromise
Users' flexibol ity in compromosong on sc:rvoce quabty attnbutes
Figure 4.4: Box Plot on Satisfaction Level with Existing OT As
In the case of satisfaction level with existing OT As, respondents who reported that
they 'may compromise' on SQAs of the airline have a median at 2 (black line). This
95
represents neutral satisfaction level and at the same time indicates 50% of the data is
greater than this value. Users' with any lesser satisfaction with existing OTAs are
represented everything above median black line, while the users with higher
satisfaction are represented everything below median black line. As shown by the top
'whisker', this group has greatest values but no outliers. Hence the data is normally
distributed.
In the case of satisfaction level with existing OT As, respondents who reported that
they 'cannot compromise' on SQAs of the airline have a median at 3 (black line). This
represents high satisfaction level and at the same time indicates 50% of the data is
greater than this value. Users' with any lesser satisfaction with existing OTAs are
represented everything above median black line, while the users with higher
satisfaction are represented everything below median black line. As shown by the top
'whisker', this group has greatest values but no outliers. Hence the data is normally
distributed.
4.2.3.2 Means Plot (H7)
The means plot is shown in Figure 4.5.
Satisfaction Level with existing Online Travel Agencies (OTAs)
I- Highly Satisfied, 5- Hi~hly Dissattsfied
2.7,----------------------,
2.6
0~ 25
19~--------------------4 Can Compronuse May Compromise Cannot Compromise
User's flexibility in compromising on semce quality attributes
Figure 4.5: Means Plot on Satisfaction Level with Existing OT As
96
The means plot shows that their apparently enormous difference between the
satisfaction level of the three respondents groups, which may appear not to be actual
case. Therefore as a follow-up and to backup this. we will analyze same results in a
different chart to see the difference between the groups.
4.2.3.3 T-Test (H7)
In this case the three groups are significantly different using a !-test (t =35.509, df =
169, p = 0.000) as shown in Table 4.6. 95% Confidence interval is probability that the
interval contains the true mean. CI of the groups is closely related to the results of the
analysis of variance for these groups. The confidence interval for each graph below
shows a linear pattern of the sample distribution which otherwise appeared to be
showing huge variations in the means plot.
Table 4.6: T-Test on User's Flexibility with Existing OTAs
Test Value= 0 -----·---::----::-::-
df Sig. (2- Mean 95% Confidence Interval of the ---~ta=iii!IJLDifference _______ _ ____ Differen=-=c:c:e __ _
t
How flexible 35.509 169 you are
4.2.3.4 Error Bars (H7)
.000 2.14 Lower
2 02 Upper 2.25
The same results are now reproduced in the error bars as shown in Figure 4.6, with
95% confidence intervals to have an idea of the variation in sample distribution.
In error bars we intend to see if the mean of one group is included in the
confidence interval of the other two groups - if so then there is likely no difference
among the groups. Moreover, it is not relevant whether the error bars 'overlap' but
whether the mean of one group 'overlaps' with the error bars of the other. The
confidence intervals can overlap by as much as 25 percent of their total length and
still show a significant difference between the means for each group. Any more
overlap and the results will not be significant.
97
Error Bar: Satisfaction Level with existing Online Travel Agencies
1- Highly Satisfied, 5- Highly Dissatisfied
3.0,-------------------------,
~ 0 28
" 0
·~ 2.6 0
" " " 2.4 .§ " .g 1l 22
·~ 2.0
, 8
'. Can Compromise May Compromise Cannot Compromise
Users' flexibility m compromising on service quality allributes
Figure 4.6: Error Bar on Satisfaction Level with Existing OTAs
In Figure 4.6, 95% CI tells us that the satisfaction level of existing OT As for users
who "can compromise" on SQAs of the airline is probably between 1.98 and 2.42,
with group mean of 2.2. Likewise, for users who "may compromise" it is probably
between 2.03 and 2.44, with group mean of 2.28, and for users who "cannot
compromise" it is probably between 2.3 and 2.8, with group mean of2.6.
The group mean of users' who 'can compromise' shares a certain degree of
confidence interval overlap with the error bars for users who 'may compromise', thus
the two groups may not necessarily be different from one another. Moreover, the
group mean of users' who 'cannot compromise' does not share any degree of
confidence interval overlap with either of the two groups, therefore, this particular
group appears to be significantly different from the rest of the sample population.
However, only with our post-hoc tests, this can be confirmed.
4.2.4 Hypothesis Testing H7
H1: The level of satisfaction with existing OTAs will be different for respondents
with different attitudes towards Users' Flexibility in compromising on SQAs of the
airline.
98
-1.2.41 One-way Analysis of Variance (if)
To test this hypothesis, one-way analysis of vanancc was used to determine
satisfaction mean of users with existing online travel agencies of the airline and at the
same time report their flexibility level in terms of compromising on SQAs of the
airline. The respondents had to select from given three options of, ( 1) Can
compromise on SQAs, (2) May compromise on SQAs, (3) Cannot compromise on
SQAs as shown in Table 4.7.
The analysis showed significant differences among satisfaction mean of the three
user groups with existing online travel agencies (F (2,169) = 6.728, p = .002 < .01).
Table 4.7: ANOVA on Satisfaction Level with Existing OTAs
Sum of df Mean F s· S S Ig . . . ___ quar~------ quare
_,B:::.:e~tw:.:.e::.:e:.::n'-'G"'r'-'o:.::uorp.-_s __ -"9_,_.4c.::2:.:_9 __ 2 __ .. _4.7_1_5 _ . __ 6._7_28 __ .00_2 __ Within Grouj!s 117.Ql8__ 167 .701 Total 126.447 169
The respondents who indicated their flexible attitude as "can compromise" on
SQAs of the airline, depicted highest level of satisfaction with existing online travel
agencies (M = 2. SO = .625). This was closely followed by satisfaction of the
respondents who indicated their flexible attitude as "may compromise" on SQAs of
the airline (M = 2.17, SO= .794). The respondents who reported their flexible attitude
as "cannot compromise" on SQAs of the airline, depicted least level of satisfaction
with existing online travel agencies (M = 2.57, SO = .984). Since the three user
groups differed significantly on satisfaction mean level of existing online travel
agencies, therefore, hypothesis H7 is accepted
-1.2.4.2 Post-hoc Sche[[e Tests (H7)
Post-hoc Scheffe tests in Table 4.8 showed that there is a significant difference
between the pair of means of the respondents who reported their flexible attitude as
"Cannot compromise" on SQAs of the airline with those who "Can compromise"; p =
.000 ( < .00 I). The same group of respondents also differed significantly from the
99
group of respondents who reported their flex1ble attitude as "May compromise" on
SQAs of the airline, p=.031 ( < .05).
Table 4.8: Multiple Comparisons on Satisfaction Level with Existing OT As
Dependent Variable: Satisfaction with Existing OTAs
fl (I~bhlow (J) how flexible D:wf ... ean Std. s· C 9fisdo/o
ext e you 1 .erenc E tg. on 1 ence you are rror
, __ are . - .... --~--~·-- ~~J)i _________ --~nterval_~ Lower Upper Bound Bound
~-----~--
Can May Compromise Compromise ______ .. ---~
-.17 .167 .579 -.59 .24
Cannot -.57(*) .166 .003 -.98 -.16
_ ~omjlromise --------------;----
May Can Compromise -.24 Comjlromise _____ ~· ---- ___ .1_7 --·~16~7~---·5_79-~- ---
.59
Cannot ____ _ Compromise
-.39(*) .148 .031 -.76 -.03 ···-~·------·---· -----
Cannot Can Compromise
Comprom_is_e_~_ .-----;----------------·--···-----::c;c--
MayCompromise .39(*) .148 .031 .03 .76
.57(*) .166 .003 .16 .98
* The mean difference is significant at the .05 level.
4.2.4.3 Effect Size for One-way ANOVA
From our hypothesis testing we know that the three groups are different, but this does
not confer the strength or the magnitude of this effect. Effect size is measure of the
strength of an effect. And since we have already rejected the null-hypothesis in the
both of the above cases, therefore it makes sense to calculate effect-size to determine
the size of the effect. The size of the effect is 7.5% (1] 2 = 0.0745).
4.3 Phase 1: Users' Satisfaction with SBTs against their rated OTA Feature
This study is to address the 3'd research question.
RQ3: How users' satisfaction with existing SBTs of airlines is rated against their
choice of OT A feature and reflected in their integration assessment of the same for
making SBTs more FOARS?
100
4.3.1 Hypothesis Testing Hs
Hs: Users' satisfaction with existing SBTs will be diflerent across their choice of four
OTA features for making SBTs more FOARS.
4.3.1.1 Means Plot
The means plot as shown in Figure 4. 7 illustrates two lines, red line indicating
respondents who consider integration of OTA features into SBTs may not necessarily
make them FOARS, while the green one denotes the respondents who think
otherwise. Means plot showed that out of the four OT A features investigated in this
study, user satisfaction was highest for opaque fare and hotel search facility. Among
the two, opaque fare was highly recommended due to being users' absolute
satisfaction point.
4.0
!i J5 ~
~
J 30
f 25
1 ] 20
6 ] ~ 15
10
M"'"'D"p]o, Op.K[ LocI ,~01 A Ite-m arc II •rport '
RecommendedOTA feature for• ntegrat10n •ntoSBT~
Do you think integration of
OTA features into existing SBTs will
make them more FOARS?
0 Notsosurc
Yes~erysure
Figure 4.7: Means Plot on the Recommendation oflntegrating OTA Features
Out of the four OTA features investigated in this study, opaque fare and hotel
search are the only two OTA features for which the green line is lower than the red
line (low value indicates high satisfaction). However, among the two, opaque fare is
the most recommended OT A feature for making SBTs flexible, since it reflects
absolute highest satisfaction point of the respondents, who thought integration of
OTA features into SBTs will make them more FOARS and also among respondents
who think otherwise, because it has been considered as the second most important
feature for integration, only after Matrix Display.
101
Table 4.9: Mean Score and SD on the Recommendation of Integrating OT A Features
Matrix Opaque Alternative Hotel
Airport Search Display Fares
Search Facility Do you think Yes. very Mean 2.25 1.60 3.00 3.40 integration of sure OTAfeatures about it. S.D. .500 .548 .707 .548 in to existing
Mean 2.00 2.06 2.83 3.67 SETs will make Not so
them more very sure S.D. .707 .827 .753 .577
FOARS? about it.
4.3.1.2 Two-ways Analysis of Variance
A two-way analysis of variance tested the satisfaction level of the representatives with
existing OARS and also reported if the integration of OTA features into SBTs would
make them more FOARS. and also picked their recommended OT A feature for
integration into SBTs. The three different F-tests as shown in Table 4.10 are:
I. The first one is the mean satisfaction level different across four proposed OT A
features for SBTs. controlling for that effect of sharing, if the chosen OTA
feature will make existing SBTs more flexible. The difference in satisfaction
level has been found to be statistically significant at p<O.Ol.
2. The second F -test looks at whether respondents who reported that integration
of OTA features into SBTs will make them flexible reservation systems, do or
do not have different levels of satisfaction with existing OARS, and again the
results were significant at p<O.OS
3. The third F -test examines the interaction effect of the four proposed OT A
features and their integration into SBTs for making them FOARS. The finding
was significant at p<O.OS, suggesting that some combination of OTA features
and existing SBTs are related and can influence upon one another, especially
in terms of making them more flexible.
102
Table 4.10: Two-way ANOV A on the Satisfaction Level with Existing OARSs
Dependent Variable: Satisfaction Level with Existing OARS (SBTs)
Source Type III Sum df _ ____ _____ ___ __ __ of Squares _
_ (_:orrectcd Model___ __ ______ 18_229(a) 7 _Intercept __ ____ ___ 269.145 I Recommended OT A feature for SBTs --·----- o--=-:--:--Flexible SBTs (Yes/No) ____ _ Flexible SBTs * Recommended OT A feature for SBTs
18.056
1.827
1.012
21.591
3
3
42 Error Total
-- -- ------ --· --·· -- ----
338.000 50 --- --···--- ---- -·--·· ----
Corrected Total 40.320 49 a R Squared~ .465 (Adjusted R Squared~ .375)
Mean .fuiuare
2.676
F
5.205 ---
Sig .
.000 269.145 523.552 .000
6.019 11.708 .000
1.827 3.556 .004
.337 .656 .042
.514
The three different F -tests in the two way analysis of variance are discussed as
under:
I. Respondents, who indicated that integration of OTA's feature into existing
SBTs will make them FOARS, shared the highest level of satisfaction when
opted for opaque fare as a recommended OTA feature for SBTs. However,
respondents who indicated integration of OTA features into existing SBTs
may not make them flexible, reported higher level of satisfaction when opted
for Matrix Display and Alternate Airport Search as a recommend solution for
SBTs (F = 11.708, p = .000 < 0.01).
2. Satisfaction level with existing OARS differed significantly (F = 3 .556, p =
.004 < 0.05) across respondents who indicated whether or not integration of
OT A features would make the existing systems more flexible.
3. The interaction effect of the four proposed OTA features and their integration
into SBTs for making them FOARS was also significant (F = .656, p = .042 <
0.05).
F-Test for respondents, who indicated that integration of OTA's feature into
existing SBTs will make them flexible Online Airline Reservation Systems, shared
the highest level of satisfaction when opted for opaque fare as a recommended OT A
feature for SBTs (F = 11.708, p = .000 < 0.01 ). Since the satisfaction level of the
users differed significantly, therefore, hypothesis is accepted
103
4.4 Phase II: Users' Perception on Flexible Tt·aveling Behavior
This study is to address the 4'11 research question.
RQ4: How users' perception on factors influencing flexible traveling behavior and
FOARS is determined?
4.4.1 Qualitative Analysis
The demographics of the 14 respondents fi:om travelblog forum and 17 from
travellerspoint forum are shown in Table 4.11. The average ages of the 14 male and
female respondents in case of Travelblog forum were recorded 37 and 35,
respectively. Likewise, the average ages of the 17 male and female respondents in
case ofTravellerspoint forum were recorded 41 and 37, respectively.
Table 4.11: Demographics of Online Travel Forums (OTF) Respondents
Male
Female
Average Age (Male)
Average Age (Female)
Countries of Origin
Rejected cases of ambiguous respondents
Travelblog (14)
9
5
37
35
USA, China, Singapore, UK, Kenya, India, Pem. Indonesia, Spain, Thailand, Nepal, Romania, Canada, South Africa
5
Travellerspoint (17)
10
7
41
37
USA, Japan, Malaysia, China, Singapore, Thailand, Australia, Indonesia
7
The responses received through online discussion forums were analyzed first as
shown in Table 4.12 in order to investigate the factors influencing on users' flexible
traveling behavior and factors influencing upon users perception of a flexible online
airline reservation system.
104
Table 4.12: Results of Online Travel Forums
Reasons/Factors ----------------------~--~--~~-----·~--~----~~-----
• Promotional packages offered by a particular
Which factors influence on users' flexible traveling behavior?
Which factors influence upon your perception of a flexible online reservation system?
airline • Searching skills to look for best offers and deals
online • Budgetary constraints and leverages • Traveling comfort in services offered • Traveling purpose • Flying frequency • Family and friends
• Simplicity • Easiness • Multiple options
After data analysis of online travel forum discussion. the semi structured in-depth
interviews were conducted; the demographics are summarized in Table 4.13.
105
Table 4.13: Demographics ofln-depth (ID) Interview
No. of respondents (as per Country Age Flying Frequency I Purchase Purchase
Gender Experience with Experience with same Country of Origin) of Origin (Avg.) Year (Avg.)
SBTs of Airline any J.lOJ.lUlar OT A
6 Malaysia Male 4 34 Frequent
Yes Yes Female 2 30 (Twice a year)
4 USA Male 2 40 Very Frequent
Yes Yes Female 2 37 (more than twice a year)
3 UK Male 2 29 Very Frequent
Yes Yes Female I 26 (more than twice a year)
---- -
2 China Male I 36 Very Frequent
Yes Yes Female I 24 (more than twice a year)
2 India Male 2 33 Frequent
Yes Yes - Female 0 - (Twice a year) 0 Male 2 43 Frequent 0'> 2 Pakistan Yes Yes
Female 0 - (Twice a vear) ' - - - -- ..1 /
2 Singapore Male 0 - Very Frequent
Yes Yes -----Female 2 27 (more than twice a year)
---·--
2 Ireland Male I 25 Frequent
Yes Yes
Female I 20 (Twice a year) -------------
2 Thailand Male I _ ll__ Frequent
Yes Yes
Female I 29 (Twice a year) ---.--·---
I Australia Male I 44 Very Frequent
Yes Yes Female 0 - (more than twice a year)
I Sudan Male I 35 Not Very Frequent
Yes Yes Female 0 - (Once a year )
I Mauritius Male 1 29 Frequent
Yes Yes Female 0 - (Twice a year)
The responses received through in-depth interviews were analyzed in the second
phase. Table 4.14 provides additional information on users' flexible traveling
behavior.
Which factors influence on users' flexible traveling behavior?
Which factors influence upon your perception of a flexible online reservation system?
Table 4.14: Results ofln-depth Interviews
Reasons
• Airlines repute • Standard conscious • Promotional packages offered by a particular airline • Searching skills to look for best offers and holiday deals • Customer loyalty • Traveling comfort in services offered • Supporter of Green Environment • Patriotism • Traveling purpose • Flying frequency • Budgetary constraints and leverages • Occupation • Children holidays • Traveling mileage
_•_Interest in recr~ation, leisure, and touri;)ll1 • Provides alternative dates for flying • Allows self-adjustments in itinerary • Simplicity • Easiness • Multi-linguistic
After data analysis of semi structured in-depth interview, focus group interviews
were conducted, the demographics are shown in Table 4.15.
Table 4.15: Demographics of Focus Group (FG) Interviews
Meeting With Independent Airline Reservation Offices
Airlines No. of No. of Executives No. of Technical
Manager interviewed Experts Malaysian Airline I 2 I Fire Fly I 2 Aero Asia I
107
Table 4.16: Results of Focus Group
Reasons ----·----. ---·
Which factors influence on users' flexible traveling behavior?
Which factors influence upon your perception of a flexible online reservation system?
• Peak I off travel time • Discounted versus Normal Fare • Traveling mileage • Traveling purpose • Working hours • Traveling comfort in services offered • Children holidays • Recreation, leisure, and tourism • Promotional Schemes • Airlin(!S repute • Easiness • Product presentation • Post sale features • User prompting for their guidance throughout reservation
process • Matrix display to sort airfares on different dates and
destinations • Low fare notifications • Flexible and alternativ<: date search • Dynamic packaging • Hotel search display, sort and reservation
The following 6 themes emerged under factors influencing upon flexible traveling
behavior and 3 themes emerged under factors influencing upon perceived flexibility
of a reservation system after giving much thought process to results, reviewing
literature, discussion with qualitative researchers:
• Theme 1: Travelers' Flexible Behavior IS molded by their traveling
consciOusness.
• Theme 2: Travelers' Flexible Behavior is molded by their belief that they
have the required digital skills.
• Theme 3: Travelers' Flexible Behavior 1s molded by their self-belief as
flexible travelers.
• Theme 4: Travelers' Flexible Behavior is molded by societal influences.
• Theme 5: Travelers' Flexible Behavior is molded by how they attribute a
cause to their traveling behavior.
• Theme 6: Travelers' Flexible Behavior is molded by their prior traveling
expenences.
108
• Theme 7: Systems Perceived Flexibility IS int1uenced by its perceived
usability.
• Theme 8: Systems Perceived Flexibility is influenced by end-user support.
• Theme 9: Systems Perceived Flexibility is influenced by comparison of
features on the actual level of effect regarding to complete the reservation
process.
4.5 Phase II: Classification of Users on the Basis of Their Flexible Traveling
Behavior
This study is to address the 51h research question.
RQ5: How to classifY Users' on the basis of their Flexible Traveling Behavior into
High, Medium and Low flexible and how to investigate interrelationships among
System's Flexibility, Users' Flexibility and Perceived Usability of existing OARS?
4.5.1 Hypotheses Testing H9
H9: Users can be classified on the basis of their flexible traveling behavior.
Users' Flexibility Transforming Scale - The transforming scoring scale was
accordingly adapted in this study as discussed in Chapter 3 to meet the requirements
of item discrimination. The same results in adapted table are shown in Table 4.17.
Table 4.17: Users' Flexibility Transforming Scoring Scale
Users' Flexibility Rating on Service Hi~hest Hi~h Neutral Low ·Least Quality Attributes of Airlines 1 2 3 4 5 Scale -2 -I 0 +I +2 Users' Rating Frequency I I 4 3 I Product (Users' Rating Frequency * -2 -1 0 0 2 Scale)
~
Users' Flexibility Score (UFS) 2
The participants scormg on Section 2 of the questionnaire (Appendix D) was
transformed using Table 4.17, in order to obtain their unique Users' Flexibility Score.
In total, ninety (90) Users' Flexibility Score was recorded and transformed, of which
109
after extensive filtration of results, 62 cases were retained of those Users' only who
had adequately respondents in all 4 sections of the questionnaire.
Range: From our sample data set, the range to classify respondents on the basis of
their Users' Flexibility Score is 2 - + 19 as shown in Table 4.18. This range has 18
digits in between inclusive of the extreme two ends. So if we divide this range
approximately among three groups, we get the £Jllowing classifications to be assigned
to users on the basis of their flexibility score.
Table 4.18: Range for Classification of Users' Flexible Behavior on the basis of their unique Users' Flexib11lity Score (UFS)
1 2 3
Classification
Low Flexible Behavior Medium Flexible Behavior High Flexible Behavior
Range
+ 02 to+ 07 + 08 to+ 13 +14to+19
Based on classification range identified from data distribution and transformation
scale adapted from literature users' classifications were made as with High, Medium
and Low flexible behavior on the basis of Users' Flexibility Score. Users' with High
Flexible Behavior were assigned code 3, Medium Flexible Behavior were assigned
code 2, and Low Flexible Behavior were assigned code 1 (Appendix E). Figure 4.8
shows the classification of users on the basis of their Users' Flexibility Scores.
-: !§
a-" 7;
6-i'
s-
·~ ,..,. I 0
------r~~ ~--~-
2 3 - '
4
Perceived Usability of EJ!:isti ng OARS
' 5
Classification of Flexible Behavior
• Lcust Flexibk Ill Medium Flt:xible D Highly Flexibk
Figure 4.8: Data Distribution aft<:r Classifications of Users'
110
Users' with high (denoted by white line), medium (denoted by green) and low
(denoted by black) t1exibility in their traveling behavior have different levels of
satisfaction. Users with high, medium and low Users' Flexibility Score have different
level of Perceived Usability and therefore the hypothesis H9 is accepted.
4.5.2 Hypotheses Testing H10
H 1o: User's Flexible Behavior and their Perceived Usability is correlated.
The interrelationship between Users' Flexible Behavior and their Perceived
Usability ofFOARS is shown in Table 4.19.
Table 4.19: Interrelationship between UFB and their Perceived Usability ofFOARS
Kendall's Tau b UFB PU Correlation Coefficient 1.000 .381(*)
Users' Flexible Behavior (UFB) Sig. (2-tailed) .049
N 62 62
Perceived Usability of FOARS Correlation Coefficient .381(*) 1.000 Sig. (2-tailed) .049
(PU) N 62 62
* Correlation is significant at the 0.05 level (2-tailed)
Negative Association- Users' Flexible Behavior (UFB) is positively associated
with Perceived Usability of FOARS, correlation coefficient r = 0.381, is significant at
p<0.05. This means that as one variable increases in value, the second variable also
increases in value. This is called a positive correlation. The significance value
indicates that the relationship is genuine, hence H1 o is confirmed.
4.5.3 Hypotheses Testing H11
H 11 : User's Flexible Behavior and System's Flexibility is correlated.
In order to investigate the interrelationship between Users' Flexible Behavior and
System's Flexibility, Pearson's Correlation Coefficients was calculated. Table 4.20
provides a matrix of correlation coefficients for the five variables, (I) Users' Flexible
Behavior, (2) System's Flexibility. (3) System's Adaptability, (4) Systems'
Ill
Adaptivity, (5) Systems' Personalization. It also displays a matrix of significance
values for these coefficients.
• Users' Flexible Behavior share a statistically non-significant relationship with
System's Flexibility, r = 0.255, p = 0.056 > 0.05.
• Users' Flexible Behavior share a statistically significant relationship with
System's Adaptability, r = 0.372, p = 0.000 < 0.001.
• Users' Flexible Behavior share a statistically non-significant relationship with
System's Adaptivity, r = 0.151, p = 0.146 > 0.05.
• Users' Flexible Behavior share a statistically significant relationship with
System's Personalization, r = 0.314, p = 0.042< 0.05.
• System's Flexibility share a statistically significant relationship with System's
Adaptability, r = 0.560, p = 0.000 < 0.001.
• System's Flexibility share a statistically significant relationship with System's
Adaptivity, r = 0.344, p = 0.045 < 0.05.
• System's Flexibility share a statistically significant relationship with System's
Personalization, r = 0.222, p = 0.032 < 0.05.
• System's Adaptability share a statistically significant relationship with
System's Adaptivity, r = 0.342, p = 0.000 < 0.001.
• System's Adaptability share a statistically significant relationship with
System's Personalization, r = 0.326, p = 0.016 < 0.05.
• System's Adaptivity share a statistically non-significant relationship with
System's Personalization, r = 0.153, p = 0.074 > 0.05.
The results showed that users' FTB to have a non-significant correlation with
System's Flexibility (r = 0.255, p = 0.056 > 0.05). Likewise, Adaptivity (sub
variables of System's Flexibility) also shared a non-significant correlation with users'
FTB (r = 0.151, p = 0.146 > 0.05). However, Adaptability (a sub-variable of System's
Flexibility) shared a strong positive and significant correlation with users' FTB (r =
0.372, p = 0.000 < 0.001). When the same results were interpreted using r2,
Adaptability accounted for 26% of the variability in users' FTB.
112
Table 4.20: Pearson Coefficient Correlations to Investigate the Interrelationship between UFB and System's Flexibility
Users' Flexible System's System's System's System's Behavior Flexibility Adaetability Adaetivity Personalization
Pearson Users' Flexible Behavior 1 0.255 0.372** 0.151 0.314* Correlation S~stem's Flexibility 0.255 1 0.560** 0.344 0.222
System' s AdaEtability 0.306** 0.560** 1 0.342** 0.326* S~stem's AdaEtivit~ 0.151 0.344 0.342** 1 0.153 System's Personalization 0.177* 0.222 0.326* 0.153 1
Sig. (2-tailed) Users' Flexible Behavior - 0.056 0.000 .146 0.042 S~stem's Flexibility 0.056 - 0.000 0.045 0.222 S~stem' s AdaEtability 0.000 0.000 - 0.000 0.016 S~stem's AdaEtivity .146 0.045 0.000 - 0.074 System's Personalization 0.042 0.222 0.016 0.074
w N Users' Flexible Behavior 61 61 61 61 61 S~stem's Flexibilit~ 61 61 61 61 61 S~stem's AdaEtabi1ity 61 61 61 61 61 System's AdaEtivity 61 61 61 61 61 System's Personalization 61 61 61 61 61
The overall relationship between UFB and SF shows non-significant values.
However, there are significant relationships between UFB and the two components of
the SF i.e. System's Adaptability and System's Personalization. Therefore, the
hypothesis H1 1 is partially accepted.
4.5.4 Hypotheses Testing H 12
H 12: Perceived Usability of OARS is not affected by users' flexible behavior after
adjusting for the effect of the covariate, System's Flexibility.
4. 5. 4.1 One-way Analysis of Variance (H12)
To meet the assumption of ANCOV A i.e. Independence of the covariate and
treatment effect - one way independent ANOV A was performed on Perceived
Usability as independent variable and covariate System's Flexibility as an outcome
variable as shown in Table 4.21. This analysis should be non-significant to meet the
assumption and result showed non-significant effect of Perceived Usability ofFOARS
on System's Flexibility.
Table 4.21: One Way Independent AN OVA with PU and System's Flexibility
Between Groups Within (Jroups Total
Sum of Squares 39.616 6.372
45.988
df 4 56 60
Mean Squa~re=-----c-F"'-_-c----=S:c:i,g,_. _ 9.904 2.476 .261 0.1138
Sum of squares between groups for the conrected model is 6.093, which indicates
total experimental effect while means square of the model is 3.047, which represents
average experimental effect as shown in Table 4.22. Unexplained variance error is the
sum of squares within groups; it is 8.128 and explains unsystematic variation within
data. The test of whether the group means are the same is represented by the F -ratio
for the combined between group effect. The value of F ratio is 22.114, which is
significant with p = .000 < 0.001. It is therefore reported after conducting ANOVA
that there was a significant effect of Users' Flexible Behavior on their Perceived
Usability ofFOARS, F (2, 59)= 22.114, p= 0.000 > 0.001.
114
Table 4.22: ANOVA with PU and System's Flexibility
Dependent Variable: Perceived Usabil ity ofFOARS
Source Type III Sum of df Mean
F Sig. S uares Square
Corrected Model __ 6.093(a) 2 3.047 22.114 .000 ----- -- --Intercept 584.054 1 584.054 4239.555 .000 - -- --Users' Flexible Behavior 6.093 2 3.047 22.114 .000 --Error 8.128 59 .138 --Total 894.345 62 Corrected Total 14.221 61
a. R Squared = .428 (Adjusted R Squared = .409)
4.5.4.2 Analysis ofCovariance (H,l)
ANOV A results indicated that an important assumption of AN COY A has not been
violated. Therefore, AN COY A was performed to first examine influence of
independent or fixed factor (Users' Flexible Behavior) on dependent variable
(Perceived Usability ofFOARS) and then experiment was manipulated by introducing
a covariate (System's Flexibility) as shown in Table 4.23 .
Table 4.23: ANCOVA by Introducing System's Flexibility as Covariate
Dependent Variable: Perceived Usability ofFOARS
Source Type III Sum df Mean
F Sig. of Squares S uare ---
Corrected Model 6.487{a2 3 2. 162 16.215 .000 InterceEt 8.400 1 8.400 62.992 .000 Users' Flexible Behavior .394 1 .394 2.952 .047 ~stem's Flexibility 3.852 2 1.926 14.443 .000 Error 7.734 58 .133 Total 894.345 62 Corrected Total 14.221 61 a. R Squared = .456 (Adjusted R Squared - .428)
Looking first at the significance value, it clear that covariate, 1.e., System's
Flexibility, significantly predicts Perceived Usability of FOARS at F( 1, 58)= 2. 952, p
= 0.47 < 0.05. What is more interesting is that when the effect of System's Flexibility
is added, the effect of Users' Flexible Behavior still remains significant (p = 0.000 <
.001) towards predicting Perceived Usability of FOARS. The amount of variation
accounted for by the model has increased to 6.487 units for the corrected model, of
which System's Flexibility now accounts for 3.9 units. Most important, the amount of
115
variation or unexplained variance in Perceived Usability of FOARS that is accounted
for by the covariate has reduced to 7.7 units from 8.1 units.
4.6 Phase III: Users' Flexibility is determined by SQAs and EVs
This study is to address the 61h research question.
RQ6: How do Service Quality Attributes of ai,rlines and External Variables jointly
predict flexible behavior oftravelers?
4.6.1 Hypothesis Testing Hn
H 13: Flexible behavior of travelers cannot be predicted by Service Quality Attributes
and External Variables.
4.6.1.1 Scatter Plots (H13)
To examine whether linear regression is appropriate, scatter plots as shown in Figure
4.9 and Figure 4.10 were examined of each independent variable against the
predicting or the dependent variable. User's Flexibility were treated as dependent
variable, SQAs and External Variables were treated as independent variable with N =
250.
~·
LJ ·s ;:s
....•.•........•
.... . .... . ......••. ... .. . .
21' .• •... . •••• !!
' 1.00
I -~-- ~-~-
2.00 3.00 ' 4.00
Service Quality Attributes of Airlines
' 5.00
R2 Linear= 0.627
Figure 4.9: Scatter Plot between Users' Flexibility and SQAs
116
The first scatter plot exammes SQAs of the airline against users' flexible
personality. Each of the points on scatter plot represents a particular observation from
the data. The data appears to be normally distributed along with linear regression line
and has no obvious outliers. A general trend or relationship between the two variables
is also predictable.
.£ :a "Q
li: ·-c " ;5
s-
·-3-
2'-
1-
I 1.00
. . . . . . . ...
I 2.00
• • r
3.00
••••
•
I 4.00
External Variable.~
I 5.00
R' Linear= 0.632
Figure 4.10: Scatter Plot between Users' Flexibility and External Variables
The second scatter plot examines external variables against users' flexible
personality. The data appears to be normally distributed along with linear regression
line and has no obvious outliers. A general trend or relationship between the two
variables is also predictable.
4. 6.1. 2 Pearson Correlation Coefficients (H13)
Correlation of the two independent variables was also computed to determine their
association with Users' Flexibility and to further ascertain their individual range and
strength of association (see Table 4.24). SQAs share a strong positive correlation with
Users' Flexibility, r = .792, highly significant at p < 0.001. Likewise, External
Variables also share a strong positive correlation with Users' Flexibility, r = .795,
highly significant at p < 0.001 level.
117
Table 4.24: Correlations between Flexible Personality, SQA and External Variables
Personality in terms SQA
External of Flexibility? Variables
How do you rate Pearson Correlation 1 .792** .795** your overall Sig. (2-tailed) .000 .000 personality in terms N 250 250 250 of flexibility? SQA Pearson Correlation .792** 1 .881**
Sig. (2-tailed) .000 .000 N 250 250 250
External Variables Pearson Correlation .795** .881 ** 1 Sig. (2-tailed) .000 .000
N 250 250 250 **Correlation is significant at the 0.01 level (2-tailed).
4. 6.1. 3 Multiple Regression Analysis (H13)
The value of Multiple Correlation Coefficient (R) between the two independent
variables and Users' Flexibility is 0.818 as shown in Table 4.25. The maximum value
of multiple correlation coefficients is 1, positive or negative and indicates correlation
of all variables for predicting one single outcome, which in this case is 0.818,
suggesting a strong relationship ofthe two independent variables with UF.
Table 4.25: Multiple Correlation Analysis between EV, SQA and Users' Flexibility
Model R R Square Std. Error of the
Adjusted R Square Estimate
I .818a .669 .667 .653 a. Predictors: (Constant), External Variables (EV), SQA
4.6.1.4 Analysis of Variance (Hn)
Analysis of Variance tests whether the model is significantly better at predicting the
outcome, than using the mean as a best guess. In this analysis, simultaneous test was
performed as shown in Table 4.26 to examine, (i) if all of the coefficient values could
be zero, and (ii) if they are all able to be zero that would mean that none of the
independent variables have a relationship with the dependent variable (null
hypothesis). And if null hypothesis is not rejected, it means the model is not useful, as
none of the independent variables have a relationship with the dependent variable.
118
In alternative hypothesis, may be at least one of the independent variable's
relationship with dependent variable will not be zero, indicating at least one of the
coefticients values is not zero. This model has an F ratio= 250.121 which is highly
significant at p <.001. This means that model significantly improves ability to
determine users' flexible behavior; therefore, null hypothesis is rejected as at least one
of the coefficient values is not zero.
Table 4.26: AN OVA with External Variables and SQAs
Model Sum of Squares df Mean Square F Sig. Regression 213.499 2 106.749 250.121 .ooo• Residual l 05.417 247 .427 Total 318.916 249 a. Predictors: (Constant), External Variables, SQAs
4.6.1.5 Testing and Interpreting Model Coefficients (H13)
Table 4.27 shows values of coefficients and !-tests.
Table 4.27: Values of Coefficients and T-test
Un-standardized Standardized Model Coefficients Coefficients t Sig.
B Std. Error Beta 1 (Constant) .119 .240 .497 .620
SQAs (Flyin£, Date Confirmation) .011 .076 .012 .143 .886 SQAs (Flyin£, Carrier Confirmation) .073 .079 .074 .930 .353 SQAs (Flyin£, Time Confirmation) -.076 .072 -.091 -1.066 .288 SQAs (No. of Stop Over). .017 .071 .018 .234 .815 SQAs (No. of connected flights) .043 .080 .042 .534 .594 SQAs (Ticket Class) .096 .064 .099 1.505 .134 SQAs (Seat Specification) .003 .068 .003 .046 .964 SQAs (Last minute discounts) .120 .066 .128 1.829 .043 SQAs (Confirmation of Origin and
.095 .061 .113 1.572 .017 Destination Airport) SQAs (Immediate Confirmation of
.106 .069 .119 1.542 .003 Itinerary on rurchase) Attribution .073 .067 .074 1.099 .273 Engagement .171 .067 .166 2.548 .011 Persuasion .071 .053 .264 1.335 .018 Identity .108 .066 .125 1.638 .037 Self Efficacy .090 .056 .070 1.619 .107
a. Dependent Variable: How do you rate your overall personality in terms of flexibility?
119
4.7 Phase III: Perceived Usability with Existing and Proposed Systems
This study is to address the 7'h research question.
RQ7: How does user Perceived Usability with the existing and the proposed system
differs?
4.7.1 Hypothesis Testing Ht4
H14: User Perceived Usability with existing and proposed systems is different across
the three groups.
4. 7.1.1 Descriptive Statistics on Effectiveness (H14)
Table 4.28 displays means, standard deviations and number of respondents in all
classifications based on flexible traveling behavior.
Table 4.28: Descriptive Statistics on Effectiveness
Dependent Variable: PU- Effectiveness
Interface Classification of Flexible Mean Std. Deviation N
Design Behavior Least Flexible 4.19 .452 30
Existing Medium Flexible 3.19 .705 19 Highly Flexible 2.96 .850 76 ..
Total 3.29 .905 125 Least Flexible 4.12 .467 30
Proposed Medium Flexible 3.74 .653 19 ----- ------Highly Flexible 4.03 .593 76 --Total 4.01 .584 125 Least Flexible 4.16 .457 60
Total Medium Flexible 3.46 .725 38 Highly Flexible 3.50 .904 152 Total 3.65 .839 250
Respondents with least flexible behavior rated the existing (mean 4.19, S.D 0.452)
and proposed systems (mean 4.12, S.D 0.467) relatively higher in terms of their
effectiveness. Likewise, respondents with a highly flexible behavior, rated the
existing systems lowest in terms of its effectiveness (mean 2.96, S.D 0.850), their
rating for the proposed systems is relatively higher in terms of its effectiveness (mean
120
4.03, S.D 0.593). These means are useful in interpreting the direction of any effects
that emerge in further analysis of variance.
4.7.1.2 Two-way Analysis of Variance on Effectiveness (1!1-1)
There is a significant effect of user's classifications on the basis of flexible traveling
behavior; since the F-ratio as shown in Table 4.29 is highly significant indicating that
the users' high, medium and low flexible traveling behavior is significantly affected
by the proposed and existing systems.
Table 4.29: Two-way ANOVA of Effectiveness
Dependent Variable: PU- Effectiveness
Source -~~f;~~a~:; ~f __§_~::e _ F Sig. Corrected Model 65.9463 5 13.189 29.410 .000 ---·-- ---·--· --· -· --· .--· --··---- ---Intercept ___ .. -- __ __ 2493.21()__ _ 2493:21§__5559.568 .OO_Q_ iJ1terfacedesig11__ ___ . ___ __ ~I. 941__ I _l_l9~ 26.627 .000 classification 20.260 2 I 0.130 22.588 .000 ---------·· ---·---- ------------· --------
interfacedesign * classificatio_11___ _H,003 _ 2 7.0Ql_ _ _l_:i.611___,000 Error I 09.423 244 .448 ------· --· --···--·---·· ·-·- ------
Total 3504.778 250 ------- -----·---·-- -- ·-
Corrected Total 175.369 249 a. R Squared~ .376 (Adjusted R Squared~ .363)
4. 7.1.3 Error Bars on Effectiveness (H14)
Figure 4.11 shows that when the influence of existing and proposed system is ignored,
the overall effectiveness of systems is very similar for users' with medium and high
flexible behavior, as the means of these two groups are approximately equaL However
the perceived effectiveness of the system for users' with least flexible behavior, not
only differs from the other two groups, it remains also higher.
121
5-
3-
~--
Least Raxible
F.rror hars: 95% Cl
Classificatlun of Flc~ibl e Bclla'o'ior
I ---
Hghly Aexible
Figure 4.11: Effect of User Classification of Flexible Behavior on Perceived Effectiveness
When the effect of existing and proposed systems is examined on perceived
effectiveness, there was a significant main effect, F (1, 244) = 26.627, p < .001. The
same data when examined in error bars as shown in Figure 4.12, the means of
proposed and existing systems were observed to be dissimilar or not equal, indicating
a probable significant relationship.
s-
J i I 1! 3"""!
' ~ .
~~I
·-! ~-·
Error hars: '.15% Cl
··-·· Proposed
1':.-isting 10.nd Propo•ed Sy•lcms
Figure 4.12: Effect of Existing and Proposed Systems on Perceived Effectiveness
4. 7.1.4lnteraction Effect on Effectiveness (H14)
In Figure 4.13, the effectiveness of the proposed system is higher for users' classified
as with medium and high flexible behavior (I ·- Least effective, 5- Highly effective).
However, for users with least flexible behavior, the effectiveness of the existing
122
system is marginally higher than the proposed one. The perceived effectiveness of the
existing and proposed system for users with least flexihle behavior also shares an
interaction effect and F -test results further reveal a signi tic ant interaction between the
effect of existing and proposed online reservation systems and the user classifications
on perceived effectiveness, F (2, 244) = 15.612, p < .001.
Pcrcciwd llsabilily of Online Airline Rcscnation Sy~tcrm (liARS)
" 0
~ 3.6-0
1 3.3"""'
d
' ~ J-
~ ' ' ' Least Re~ib~ Medium FJe:tible Highly Fle>Cible
Classifi~adon ofFinlhl~ Bduvior
OARS
h,,\in~
l',oro«:d
Figure 4.13: Interaction Effect of Existing and Proposed Systems on Perceived
Effectiveness
4. 7.1.5 Descriptive Statistics on Efficiency (H14)
Table 4.30 displays means, standard deviations and number of respondents in all
classifications based on flexible traveling behavior.
Table 4.30: Descriptive Statistics on Efficiency
Dependent Variable: PU- Efficiency
Interface Classification of Flexible Mean
Design --- .. -
Existing
Proposed
Total
Behavior -··-- --
Least Flexible Medium Flexible Highly Flexibl~. Total Least Flexible Mediwn Flexible
·- -----
--·-· -··-·-
Highly Flexibitc_ ··- _ Total Least Flexible
4.19 3.79 3.71 3.83 4.01 3.84 4.00 3.98 4.10
Medium Flexible 3.82 --.-- .-··.-·
Highly Flexible_. --··-· _ 3.85 Total 3.91
123
Std. Deviation ·---
.358 -··--
.298 ·-·--
.558 -------
.521
.514
.661
.473 - ·-
.514 .-.-
.448
N
30 19
------
76 125 30 19 76 125 60
·-·- .-----
.506 38 ------ ----
.535 152 --
.52! 250
Respondents with least flexible behavior rated the existing (mean 4.19, S.D 0.358)
and proposed systems (mean 4.01, S.D 0.514) relatively higher in terms of their
efficiency. Likewise, respondents with a highly flexible behavior, although rated the
existing systems lowest in terms of its efficiency (mean 3.71, S.D 0.558), their rating
for the proposed systems is relatively higher in terms of its efficiency (mean 4.00, S.D
0.4 73 ). These means will be useful in interpreting the direction of any effects that
emerge in further analysis of variance.
4. 7.1. 6 Two-way Analysis of Variance on Efficiency (H14)
There is a significant effect of user's classifications on the basis of flexible traveling
behavior; since the F-ratio as shown in Table 4. 31 is highly significant indicating that
the users' high, medium and low flexible traveling behavior is significantly affected
by the proposed and existing systems.
Table 4.31: Two-way ANOVA of Efficiency Dependent Variable: PU- Efficiency
Source Type III Sum
df Mean
ofSguares Sguare Corrected Model 6.714" 5 1.343 Intercept 2793.594 1 2793.594 interfacedesign .136 1 .136
·----· classification 3.029 2 1.515 interfacedesign * classification 2.436 2 1.218 ..
Error 60.935 244 .250 -Total 3880.556 250 ·--·--
Corrected Total 67.648 249 a. R Squared- .099 (Adjusted R Squared- .081)
4. 7.1. 7 Error Bars on Efficiency (Hu)
F Sig.
5.377 .000 11186.352 .000
.545 .461 6.065 .003 4.878 .008
Figure 4.14 shows that when the influence of existing and proposed system is ignored,
the overall efficiency of systems is very similar for users' with medium and high
flexible behavior, as the means of these two groups are approximately equal. However
the perceived efficiency of the system for users' with least flexible behavior, not only
differs from the other two groups, it remains also higher.
124
s- Frror bars: 95'/o Cl
o-- I I _~._ --- I
Least Flexible Madlum Flexible Hghly Flexible
ClauiOution of flotiblo Boha>lor
Figure 4.14: Effect of User Classification of Flexible Behavior on Perceived
Efficiency
When the effect of existing and proposed systems is examined on perceived
efficiency, there was not a significant main effect, F (1, 244) = 0.545, p = .461. The
same data when examined in error bars Figure 4.15, the means of proposed and
existing systems were observed to be similar or equal, indicating a probable
insignificant relationship.
5~ &ror bon: 95"1. Cl
I E>usbng
I Ropoeed
Exlsllng and Propoud Sylltms
Figure 4.15: Effect of Existing and Proposed Systems on Perceived Efficiency
-1. 7. 1.8 Interaction Effect on Efficiency (Hf.l)
In Figure 4.16, the efficiency of the proposed system is higher for users' classified as
with medium and high flexible behavior (1 -Least effective, 5- Highly effective).
125
However, for users with least flexible behavior, the efficiency of the existing system
is marginally higher than the proposed one. The perceived efficiency of the existing
and proposed system for users with least flexible behavior also shares an interaction
effect and F -test results further reveal a significant interaction between the effect of
existing and proposed online reservation systems and the user classifications on
perceived efficiency, F (2, 244) = 4.878, p < .05.
Perceived Usability of Online Airline ReselrVAtion Systems (OARS)
~ 8. ' .J: 4.1-
~ ~ ..
~ 4_1 ~ " 'I. ·1: 3.9""1 t . u i ;>-. il
~ 3.a-: § I'
- i 3.7""]
·--~- ·-r·---~~~-- r-~~~--~~---r
Least Aexible !'tedium Flexible Hghly Flextble
Classlflcalion ofF1ulble Behavior
OAR~
-Existing ----- Prllpo•~d
Figure 4.16: Interaction Effect of Existing and Proposed Systems on Perceived Efficiency
4. 7.1. 9 Descriptive Statistics on Satisfaction (H14)
Table 4.32 displays means, standard deviations and number of respondents in all
classifications based on flexible traveling behavior.
Respondents with least flexible behavior rated the existing (mean 4.10, S.D 0.377)
and proposed system (mean 4.04, S.D 0.393) relatively higher in terms of their
satisfaction, and their satisfaction has also fallen by 6% with proposed systems.
Likewise, respondents with a highly flexible behavior, although rated the existing
systems lowest in terms of satisfaction (mean 3.44, S.D 0.576), their rating for the
proposed systems is relatively higher in terms of satisfaction (mean 3.68, S.D 0.556).
These means will be useful in interpreting the direction of any effects that emerge in
further analysis of variance.
126
Table 4.32: Descriptive Statistics of Satisfaction
Dependent Variable: PU -Satisfaction
Interface Classification of Flexible _!)_e~ign .- _ _lleltavior
Existing
Proposed
Total
Least Flexible -----
Medium Flexible -- .. - --·
Highly Flexi!Jle Total Least Flexible Medium Flexible ----·-··-
Highly Flexible __ Total Least Flexible
-·-
Medium Flexible ---- ----- --
Highly Flexible __ . _. __ Total
4. 7.1.10 Two-way Analysis of Variance (H14)
Mean
4.10 -
3.60 3.44 3.62 4.04 3.80 3.92 3.93 4.07 3.70 3.68 3.78
Std. N
Deviation -···-· - -
.377 30 ·- -·
.348 19 - --.-
.576 76 - -
.573 125
.393 30 --·· -· .-- --·· .-
.582 19 --------
.417 76
.443 125 -·.- .. -· -----
.383 60 -------- ------- --··
.484 38
.556 152 -· -- -- ----
.534 250
There is a significant effect of user's classifications on the basis of flexible traveling
behavior; since the F-ratio is highly significant as shown in Table 4.33 indicating that
the users' high, medium and low flexible traveling behavior is significantly affected
by the proposed and existing systems.
Table 4.33: Two-way ANOV A of Satisfaction
Dependent Variable: PU- Satisfaction
Source Type III Sum df Mean F Sig. ______ _ __ of Squares ____ Square Corrected Mod-:-e-:-1-- 16.2303 5 3.246 14.454 .000 -----· ---- --· ·-·· -· --··--- ----·--·-··--
Intercept ______ 2_6_46.1_9_1 ___ 1 _26_4_6._19_1_1_17_8_2:88_1 :00_0 interfacedesign___ 1_.960_ I _1_.96_0 ___ 8_.728 .0_0_3 classification 6.952 2 3.476 15.477 .000 ------ ---------- - ------ - -
interfacedesign *classification .- 3.31 I__ 2 _L6_56_ _ _7.372 .001 Error 54.797 244 .225 - -- --- --------- -- ----- ------
Total 3636.516 250 ---- ------
Corrected Total 71.028 249 a. R Squared~ .229 (Adjusted R Squared~ .213)
4. 7.1.11 Error Bars on Satisfaction (H14)
Figure 4.17 represent the error bars on satisfaction.
127
When the influence of existing and proposed system is ignored, the overall
satisfaction of systems is very similar for users' with medium and high flexible
behavior, as the means of these two groups are approximately equal. However the
perceived satisfaction of the system for users' with least flexible behavior, not only
differs from the other two groups, it remains also higher.
s-~ Frror bors : 95•1. a
~ ·- ~
I -.--L.
c 0
-=~
~ "' : ~2-
1-
~ I '~ I 1 - ~
l.MitAextJio Mocfumfltxlble HgNyAoo><tJio
Cl•u lncafion offln1bh: B~hl\'lor
Figure 4.17: Effect of User Classification ofFlexible Behavior on Perceived
Satisfaction
When the effect of existing and proposed systems is examined on perceived
satisfaction, there was a significant main effect, F (1, 244) = 8.728, p < .05. The same
data when examined in error bars Figure 4.18, the means of proposed and existing
systems were observed to be dissimilar or not equal, indicating a probable significant
relationship.
. 0
~3-"' . . i2-
Frror bon: 95% Cl
--~-- --~-
E'lcabng
~-"'
Figure 4.18: Effect of Existing and Proposed Systems on Perceived Satisfaction
128
4. 7.1.12 Interaction Effect on Satisfaction (Hu)
In Figure 4.19, the satisfaction of the proposed system is higher for users' classified as
with medium and high flexible behavior (I - Least effective, 5- Highly effective).
However, for users with least flexible behavior, the satisfaction of the existing system
is marginally higher than the proposed one. The perceived efficiency of the existing
and proposed system for users with least flexible behavior also shares an interaction
effect and F -test results further reveal a significant interaction between the effect of
existing and proposed online reservation systems and the user classifications on
perceived satisfaction, F (2, 244) = 7.372, p < .05.
Perceived li!iahility of Online Airline Rcscnation Systems (OARS)
' ' ' Least Aex1ble r.kdium Flexible Highly Rexible
Classilication of Flexible Bch a~or
OARS
-!:.XJ>llllg
PrupuseJ
Figure 4.19: Interaction Effect of Existing and Proposed Systems on Perceived Satisfaction
4. 7.1. I 3 Levene's Test on Existing Systems (HJo)
Results of Levene's Test as shown in Table 4.34 show non-significant results
(p=.l40>.05) indicating homogeneity of variance assumption being met, therefore,
post hoc analysis can be performed by using Scheffe Test.
Table 4.34: Levene's Test on Existing Systems
Dependent Variable: Perceived Usability (Effectiveness+Efficiency+Satisfaction)
F dfl df2 Sig. 1.995 2 122 .140
Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept+ classification+ interfacedesign +classification* interfacedesign
129
4. 7. 1.14 Post-hoc Multiple Comparisons on Exi;,ting Systems (H14)
Post-hoc Multiple Comparisons using Scheffe Test is shown in Table 4.35. In case of
existing systems, users' classified as with High, Medium and Low flexible behavior
did not differ from one another in terms of rating Perceived Usability of the existing
OARS.
Table 4.35: Post-hoc Scheffe Multiple Comparison Using Schefe to Test Users' Classification for Existing Systems
Dependent Variable: Perceived Usability (Effectiven.,ss, Efficiency, Satisfaction) of Existing System_s_ __
(I) Classification (J) Classification Mean 95% Confidence
Std. Interval of Flexible of Flexible Difference
Error Sig.
Lower Upper Behavior Behavior (1-J)
Bound Bound --·---~-- -------
Least Flexible Medium Flexible .2645 .13027 .132 -.0583 .5873
--------- ·- . -----Highly Flexible .0779 .09580 .719 -.1595 .3153
--------· -- -------
Medium Flexible Least Flexible -.2645 .13027 .132 -.5873 .0583 -------Highly Flexible -.1866 .11396 .266 -.4690 .0958
-~--------
Highly Flexible Least Flexible -.0779 .09580 .719 -.3153 .1595 --~-~--- . ·--Medium Flexible .1866 .11396 .266 -.0958 .4690
Based on observed means. The error term is Mean Square (Error)~ .197.
4. 7.1. 15 Levene's Test on Proposed Systems (H 14)
Results of Levene's Test as shown in Table 4.36 show non-significant results (p =
.112 > .05) indicating homogeneity of variance assumption being met, therefore, post
hoc analysis can be performed by using Scheffe Test.
Table 4.36: Levene's Test on Proposed Systems
Dependent Variable:PU4
F dfl df2 Sig. 5.799 2 122 .112
Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept+ classification+ interfacedesign --classification * interfacedesign
4. 7.1.16 Post-hoc Multiple Comparisons on Proposed Systems (H14)
Table 4.37 shows post hoc multiple comparison of Users' classification on the basis
of their flexible traveling behavior. In rating Perceived Usability of the proposed
FOARS, Users' classified as with Least Flexible in their traveling behavior differed
130
significantly from Users' classified as with Medium Flexible (at p < .001) and Highly
Flexible (at p < .001) and vice versa.
Table 4.37: Post-hoc Multiple Comparison Using to Schefte Test Users'
Classification for Proposed Systems
Dependent Variable: Perceived Usability (Effectiveness. Et1iciency. Satisfaction) of Proposed Systems
(I) 95% Confidence Classification (J) Classification Mean
of Flexible Difference of Flexible
Std.
Behavior Error (1-J) Behavior
-~-···--···--·~~~-
. Medium Flexible .6336* Least Flextble . HighlYFl~xible - .... 7911 *
Interval Sig. U Lower pper
Bound Bound =-:c~
.000 .2637 1.0035
.000 .5191 1.0631 ··-- ~---
Medium Least Flexible -.6336* .000 -1.0035 -.2637 - .--·· ·-- ·--- ----- -~----·~--·-~-
.485 -.1661 .4811 Flexib_Jt:_ _Highly Flexible ___ .1575 .
.14926
.10977
.14926
.13058
.10977
.13058
- ~- --- -~~
Hi hi Flexible Least Flexib_le _____ -.791_1 *_ g y Medium Flexible -.1575
Based on observed means. The error term is Mean Square(Error) ~ .259. *.The mean difference is significant at the .05 level.
.000 -1.0631 -.5191
.485 -.4811 .1661
4.8 Phase III: Effect of Users' Flexibility on Proposed Systems
This study is to address the 81h research question.
RQ8: Is there a multivariate main effect of user's Flexible Traveling Behavior (High,
Medium and Low) on effectiveness, efficiency and satisfaction of the proposed
system?
4.8.1 Hypothesis Testing Hts
H 15: There are differences among effectiveness, efficiency and satisfaction caused by
the users' Flexible Traveling Behavior.
4.8.1.1 Scatter Plots (His)
Pair wise nonlinear relationships between dependent variables using scatter plots are
shown in Figure 4.20, 4.21 and 4.22. In Effectiveness Versus Efficiency,
Effectiveness Versus Satisfaction, Efficiency Versus Satisfaction, Strong, positive,
131
linear relationships is observed as one variable increases in value, the other variable
tends to also increase.
s-·
"-,_
~ :8 J.s-
IIi "' e-
2.5-,
I
! ,_.
Cla~silication ofAniblc
/. "'""""' Otc.J<t Flcxobk OMcdournFlnoblc
lhghly flcx1bk
• R' [_mcar~tu37
--I ----.------I ---r- I -T-2_5 3 3.5 4 4.5 5
Efficiency
Figure 4.20: Linear Relationship between Effectiveness and Efficiency
25
I
"L,-. ,,
'/ Clauification of Flexible • Behavior
8Lo~.« Flcx•bk M~diwu l'l~x1ble
• Highly Flexible
R' Lmca1 - 0.4~6
- - r-~~----~--- --~r-------,--
3 3.5 4 4 5 5
Satisfaction
Figure 4.21: Linear Relationship between Effectiveness and Satisfaction
' s;
4.5...;1
I
! ·~ t
~
~ ,,ll .•
s ' "' ' ' ' ,~
,,~
• Classifi~ation
/
nfflexlblc • Behavior
8Leasl t'lc"hk Medium FleXIble l11~hly Fkxoble
R' L1ncar- U 4)4
,--~- ---. ~--~~~- -, - ---r-~-~-- ., I
2 s 3 J.s 4 .ts s
Satistaction
Figure 4.22: Linear Relationship between Efficiency and Satisfaction
132
4.8. 1.2 Homogeneity of Covariance's (!115)
The second assumption of multivariate analysis was met by examimng Box's M,
which tests the hypothesis that the covariance matrices of the dependent variables are
significantly different across levels of the independent variable as show in Table 4.38.
Table 4.38: Box's Test of Equality of Covariance Matrices
Box'sM F
--------
dfl
df2 -- ------Si .
16.935 1.337
-----
12 14037.348 ------
.189 Tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups.
Results showed non-significant results (p = .189 > .00 I) hence indicating that the
assumption has not been violated.
The overall F test for the three dependent variables was examined in Multivariate
Tests as shown in Table 4.39 by analyzing the statistic called Wilks' lambda (A.), and
the F value associated with that. In the case of Independent Variable (IV), User
classifications on the basis of their flexible traveling behavior, Wilks' lambda is .667,
and has an associated F of 8.971, which is significant at p <001. Furthermore, the
partial eta squared (partial E2) associated with the main effect of Users' classification
is .183 and the power to detect the main effect is I. Thus, H15 was accepted.
Initial interpretation of results based on one-way MANOV A have revealed a
significant multivariate main effect for User' classifications, Wilks' A. = .667, F (6,
240.000) = 8.971, p <. 001, partial E2 = .183. Power to detect the effect was I. Thus,
H 15 was accepted due to statistically significant impact of Users' classification on
three dependent variables measuring users' Perceived Usability of the proposed
FOARS. Since, the results for hypothesis testing were statistically significant, so
follow-up tests were performed and interpreted.
133
Table 4.39: Multivariate Tests
Effect Value F Hypothesis Error
Sig. Partial Eta Non cent. Observed
df df Squared Parameter Powerb -~~~
~--··-
Intercept Pillai's Trace .982 2188.873a 3.000 120.000 .000 .982 6566.620 1.000 -----~ .- ·--~~.~
Wilks' Lambda .018 2188.873a 3.000 120.000 .000 .982 6566.620 1.000 Hotelling's Trace 54.722 2188.873a 3.000 120.000 .000 .982 6566.620 1.000 Roy's Largest Root 54.722 2188.873a 3.000 120.000 .000 .982 6566.620 1.000
Classification Pillai's Trac e .333 8.069 6.000 242.000 .000 .167 48.416 1.000 Wilks' Lambda .667 8.971a 6.000 240.000 .000 .183 53.828 1.000 Hotelling's Trace .498 9.876 6.000 238.000 .000 .199 59.259 1.000 Roy's Largest Root .496 20.013c 3.000 121.000 .000 .332 60.038 1.000
a. Exact statistic
w b. Computed using alpha~ .05 ..,. c. The statistic is an upper bound on F that yields a lower bound on the significance level. d. Design: Intercept+ classification
4. 8.1. 3 Alpha Adjustment ( H15)
The experiment -wise alpha protection provided by the overall or omnibus F test does
not extend to the univariate tests. It is thus important to make an alpha adjustment to
account for multiple ANOV As being run. Hence, confidence level is divided by the
number of tests to be performed. as in this case, F tests for the three dependent
variables is required to be at p < 0.017 (.05/3).
4.8.1.4 Univariate ANOVAs (H15)
Table 4.40 shows that Users' classification on the basis of their flexible traveling
behavior have a statistically significant effect on three dependent variables assessing
Perceived Usability of FOARS, Effectiveness (F (2, 122) = 28.680; p = .000 < .017;
partial s2 = .320), Efficiency (F (2, 122) = 10,776; p = .000 < .017; partial s2 = .15)
and Satisfaction (F (2, 122) = 18.738; p = .000 < .017; partial s2 = .235).
135
Table 4.40: Univariate ANOV As
Source . Type III Sum df Mean
F Sig. Partial Eta Noncent. Observed
Dependent Vanable f S Sguare Sguared Parameter Powerb o guares Corrected Model PU - Effectiveness 32.449a 2 16.225 28.680 .000 .320 57.360 1.000
PU - Efficiency 5.058c 2 2.529 10.776 .000 .150 21.551 .989 PU - Satisfaction 9.572d 2 4.786 18.738 .000 .235 37.475 1.000
~--~-
Intercept PU - Effectiveness 1080.033 1 1080.033 1909.144 .000 .940 1909.144 1.000 ·~--~----
PU - Efficiency 1377.358 1 1377.358 5868.105 .000 .980 5868.105 1.000 PU - Satisfaction 1252.057 1 1252.057 4901.941 .000 .976 4901.941 1.000
classification PU - Effectiveness 32.449 2 16.225 28.680 .000 .320 57.360 1.000 PU - Efficiency 5.058 2 2.529 10.776 .000 .150 21.551 .989 PU - Satisfaction 9.572 2 4.786 18.738 .000 .235 37.475 1.000
Error - PU - Effectiveness 69.017 122 .566 w PU - Efficiency 28.636 122 .235 "' PU - Satisfaction 31.161 122 .255
-· ---
Total PU - Effectiveness 1457.222 125 PU - Efficiency 1871.778 125 PU - Satisfaction 1680.594 125
Corrected Total PU - Effectiveness 101.467 124 ------
PU - Efficiency 33.694 124 PU - Satisfaction 40.733 124
a. R Squared- .320 (Adjusted R Squared- .309) b. Computed using alpha = .05 c. R Squared= .150 (Adjusted R Squared= . 136) d. R Squared= .235 (Adjusted R Squared= .222)
4.8.2 Hypothesis Testing H16
H16: Effectiveness, efficiency and satisfaction in the proposed FOARS is highest for
users with highest flexible behavior.
4.8.2.1 Descriptive Statistics on Perceived Usability (H16)
The descriptive analysis on Effectiveness, Efficiency and Satisfaction is presented in
Table 4.41 and bar charts are shown in Figure 4.23, 4.24 and 4.25, respectively.
Table 4.41: Descriptive Statistics on PU
Classification of Flexible Behavior Mean Std. Deviation N -·-· ·-- -- ----- --- ---------
PU - Effectiveness Highly Flexible __ 4.16 .457 60 ·-···--·--··-··
Medium Flexible 3.46 .725 38 ---------- .-··-
Least Flexible 3.50 .904 !52 ----- ----·---- .- ·-·--·----
Total 3.65 .839 250 .- --- ·------ - ---- ·- ·-· -----·--··-···-
4.10 PU - Efficiency Highly Flexible_ .448 60 ---- --·.- --
Medium Flexible 3.82 .506 38 ·------- ------·-
Least Flexible 3.85 .535 152 - - ---- --· -- ·---- ·-- --·
Total 3.91 .521 250 ·-- ----·--·-·- ------ -- --·--·--·-
PU- Satisfaction !fighly Flexible _4.07_ .383 60 -· --.. --- -----·--
Medium Flexible 3.70 .484 38 -·· -·· ---- -- --· -- - --- -----· ---·-
Least Flexible 3.68 .556 152 -··- - ·- ·-·-- --·-
Total 3.78 .534 250
Classification
.MedJum Flextble
!-----~~-~-~---~---.! .• Least flexoble 20 25 30 _j<; 40 45 'ill
Effed tveness of I he Proposed F OARS
Figure 4.23: Bar Chart Showing Effectiveness of the Proposed System
137
Classification
01-iighly Flexible
IIJIMedium Flexible
!----~--~--~--~·--~----,! .Least Flexible J 2 34 J. 6 38 4_0 4_2 44
Efficiency oft he Proposed FOA RS
Figure 4.24: Bar Chart Showing Efficiency of the Proposed System
Classification
OH1ghly Flexible
.Medium Flexible
lr----~---~----~----1-Least Fle:>:ib!e 3.0 3.5 40 45 5 0
Satisfaction of the ProposedFOARS
Figure 4.25: Bar Chart Showing Satisfaction of the Proposed System
4. 8. 2.2 Levene's Test of Equality of Error Variances (HJ6)
Homogeneity of variances for the three dependent variables with significant
Univeriate ANOVAs was examined as shown in Table 4.42. The Levene's statistics
for the three DV s showed non-significant results (Effectiveness; p = .40 I > .05,
Efficiency; p =.051 > 0.05, Satisfaction; p = .200 > 0.05). This indicated that the
138
group variances were equal and post-hoc comparison of pair wise group means could
be examined by computing Sheffe test and (}ames-Howell.
Table 4.42: Levene's Test of Equality of hror Variances
F dfl df2 _Sig .. _ PU - Effectiveness .921 2 122 .401
------- ----
PU -Efficiency 3.048 2 122 .051 --
PU - Satisfaction 1.631 2 122 .200 .-··- -- ---
Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept+ users' classification
4.8.2.3 Alpha Correctionsfor Post-hoc Multiple Comparisons (H16)
Since Post Hoc multiple comparisons involve 9 tests to be performed therefore,
confidence level has been reset at .05/9=.005. Post Hoc multiple comparisons in terms
of means scores are shown in Table 4.43.
4.82.4 Post-hoc Multiple Comparisons (H1 6)
Post-hoc Multiple Comparisons using Scheffe Test is shown in Table 4.43. In case of
existing systems, users' classified as with High, Medium and Low flexible behavior
did not differ trom one another in terms of rating Perceived Usability of the existing
OARS.
Effectiveness - in terms of effectiveness of the proposed system, users with
highly flexible behavior differed significantly from users with medium (p=.000<.005)
and least flexible (p=.000<.005) behavior. This is also evident from descriptive
statistics table that users with highly t1exible behavior rated the proposed systems
effectiveness highest (M=4.!6, SO =.457) which is way above the average rating by
other groups.
Efficiency - in terms of efficiency of the proposed system, users with highly
flexible behavior differed significantly from users with least t1exible (p=.000<.005)
behavior. This is also evident from descriptive statistics table that users with highly
flexible behavior rated the proposed systems efficiency highest (M=4.1 0, SO =.448).
139
Satisfaction - in terms of satisfaction of the proposed system, users with highly
flexible behavior differed significantly from us<~rs with medium (p=.004<.005) and
least flexible (p=.000<.005) behavior. This is also evident from descriptive statistics
table that users with highly flexible behavior rated the proposed systems satisfaction
highest (M=4.07, SD =.383) which is way above the average rating by other groups.
Based on these results H 16 is accepted since users with highly flexible behavior
differed significantly from users with medium and low flexible behavior in terms of
rating effectiveness, efficiency and satisfaction of the proposed system.
140
Table 4.43: Post-hoc Multiple Comparisons using Scheffe
(I) Classification of (J) Classification of Mean Std. 95% Confidence Interval --- -------
Dependent Variable Flexible Behavior Flexible Behavior Difference (1-J) Error
Sig. Lower Upper Bound Bound
--- -- ------ ------ ------ --- --------
PU- Scheffe Highly Flexible
Medium Flexible 1.00* .221 .000 .45 1.54 -~-·"---- -
Effectiveness Least Flexible 1.22* .162 .000 .82 1.63 -------------- ---- ------------- --- --- ----
Medium Flexible Least Flexible -1.00* .221 .000 -1.54 -.45
----
Highly Flexible .23 .193 .499 -.25 .71 ----- --- -----
Least Flexible Highly Flexible _____ :1.22* .162 .000 -1.63 -.82
-------- --
Medium Flexible -.23 .193 .499 -.71 .25 -- ---- - ---- ------ -
Games-Highly Flexible
Medium Flexible 1.00* .182 .000 .55 1.45 -- --
Howell Least Flexible 1.22* .128 .000 .92 1.53 ------ ----
Medium Flexible Least Flexible -1.00* .182 .000 -1.45 -.55
- ---- ----
-1'>- Highly Flexible _ .23 .189 .458 -.24 .69 --- ---
Least Flexible Highly Flexible ____ -1.22* .128 .000 -1.53 -.92
- ----- -
Medium Flexible -.23 .189 .458 -.69 .24 ------ ----- - -- ---- -----
PU- Efficiency Scheffe Highly Flexible
Medium Flexible .40* .142 .022 .05 .75 --- -------- ------ -
Least Flexible .48* .1 04 .000 .22 .74 ------ ----- ----- -------
Medium Flexible Least Flexible -.40* .142 .022 -.75 -.05
- --- ---
___ Highly Flexi[)k_ .08 .124 .799 -.22 .39 --- - ------- ----- ---- -------- ------
Least Flexible Highly Fle_xible _ -.48* .104 .000 -.74 -.22
--------- --- ---- --------- -
Medium Flexible -.08 .124 .799 -.39 .22 - --- -- --- ---- ---- ---- ---- --
Games-Highly Flexible
Medium Flexible .40* .095 .000 .17 .63 ------ - ------ --- -------
Howell Least Flexible .48* .091 .000 .26 .70 -- --- ----- -------·
Medium Flexible Least Flexible -.40* .095 .000 -.63 -.17
-------- ---
Highly Flexible .08 .094 .649 -.14 .31 -- --- ----------- ---- -------------- ---
Least Flexible Hig_hly Flexibl_e __ -.48* .091 .000 -.70 -.26
- ---- ------- - ---
Medium Flexible -.08 .094 .649 -.31 .14 ---
--1'-N
Table 4.43: Post-hoc Multiple Comparisons using Scheffe (continue)
Dependent Variable
PU - Satisfaction Scheffe
GamesHowell
Based on observed means.
(I) Classification of Flexible Behavior
Highly Flexible
Medium Flexible
Least Flexible
Highly Flexible
Medium Flexible
Least Flexible
The error term is Mean Square(Error) ~ .255. * The mean difference is significant at the .05 level.
(J) Classification of Mean Std. s· Flexible Behavior Difference (I-J) Error Ig.
Medium Flexible .51* .148 .004 Least Flexible .67* .109 .000 Least Flexible -.51* .148 .004 Highly Flexible .16 .130 .464 Highly Flexible -.67* .109 .000 Medium Flexible -.16 .130 .464 Medium Flexible .51* .105 .000 Least Flexible .67* .095 .000 Least Flexible -.51* .105 .000 Highly Flexible .16 .104 .275 Highly Flexible -.67* .095 .000 iviediutn Flexible -.16 .104 .275
95% Confidence Interval ~~~--
Lower Upper Bound Bound
.14 .87
.40 .94 -.87 -.14 -.16 .48 -.94 -.40 -.48 .16 .25 .76 .44 .89
~-
-.76 -.25 ~-~-
-.09 .41 ~·-~-
-.89 -.44 -.41 .09
4.9 Chapter Summary
In this chapter the results obtained from the studies and case study are presented. To
achieve this goal. this chapter followed the research objectives as described earlier in
Chapter 3 with corresponding hypothesis to organize the results. The results obtained
from the studies helped identified assessing user needs in terms of System's
Flexibility and Users' Flexibility. This was followed by the classification of users on
the basis of their flexible traveling behavior. To validate the proposed framework for
the development of flexible online airline reservations systems, a case study was
conducted. The case study was conducted using quantitative technique whereby
participants were requested to report their effectiveness. efficiency and satisfaction
with the proposed systems using prototype. The results are discussed in detail in
Chapter 5.
143
CHAPTER 5
DISCUSSION
5.0 Chapter Overview
This chapter discusses the results obtained in chapter 4 based on five different studies
and a case study. The overall research was based on investigating three core research
objectives with eight corresponding research questions and sixteen corresponding
hypothesis. Results obtained during the study supported most of the hypotheses as
shown in Table 5.1. The following sections will discuss these results in detail.
To achieve the I st research objective. Section 5.1, 5.2 and 5.3 are dedicated for the
discussion on the results obtained from the three studies conducted in Phase I
answering the corresponding research questions RQl, RQ2 and RQ3, respectively. To
attain the 2"d research objective, Section 5.4 and 5.5 present discussions on the results
obtained from the two studies conducted in Phase II answering the corresponding
research questions RQ4 and RQ5, respectively. To conquer the 3'd research objective,
Section 5.6, 5.7 and 5.8 are devoted for the discussion on the results obtained from the
case study answering the corresponding research questions RQ6, RQ7 and RQ8,
respectively.
Finally, Section 5.9 discusses the recommendations for the designing of flexible
Online Airline Reservation Systems and Section 5.10 cover the chapter summary.
--1'-V>
Research Objectives
To assess user needs (System's Flexibility and Users' Flexibility) associated with Online Airline Reservation Systems.
Table 5.1: Summary of Research Questions, Hypotheses and Results
Research Questions Hypotheses Results
RQI: What are the issues with H 1: Non-functional Requirements are perceived to have an S rt d flexibility of Online Airline impact on the usability of OARS. uppo e Reservation Systems, whether or not H2: Functional Requirements are perceived to have an Not flexibility is one of the reasons for impact on the usability of OARS. Supported users not using such systems? H3: The perceived flexibility of OARS affects the S rt d
usability of such systems. uppo e
RQ2: To what extend flexible users can compromise with service quaiity attributes of Online Airline Reservation Systems?
H4: Functional Requirements of OARS are inversely associated with the flexibility of thesystems. H 5: The availability of resources and skills set influence upon the usability of OARS. H 6: The level of satisfaction with existing SBTs is different for respondents \Vith different attitudes to,vards Users' Flexibility in compromising on SQAs of the airline.
H 7: The level of satisfaction with existing OT As is different for respondents with different attitudes towards Users' Flexibility in compromising on SQAs of the airline.
RQ3: How users' satisfaction with an H8: Users' satisfaction with existing SBTs is different existing SBTs is rated against their across their choice of four OTA features for making SBTs choice of OTA feature and reflected in more FOARS. their integration assessment of the same for making SBTs more flexible Online Airline Reservation Systems?
Supported
Supported
Supported
Supported
Supported
... 0\
Research Objectives
To investigate users' perception on factors influencing Flexible Traveling Behavior and to classify them on the basis of their Flexible Traveling Behavior.
To study the relationship between users' Perceived Usability and travelers Flexible Traveling Behavior with existing and proposed Online Airline Reservation Systems.
Table 5.1: Summary of Research Questions, Hypotheses and Results (continue)
Research Questions
RQ4: How users' perception on factors influencing flexible traveling behavior and flexible OARS is determined?
Hypotheses
RQS: How to classify Users' on the H 9: Users can be classified on the basis of their Flexible
Results
Supported
basis of their Fl exib I e Traveling _T~ra=-vc.:eo:l:.:in'-'g'-oB=-::eh:.;-a=-v.:..:i:;:o-;-r.'--:::-:--:-----,--;--,---,;::--:----;---------Behavior into High, Medium and Low H 10 : User's Flexible Behavior and their Perceived S rt d
Supported
flexible and how to investigate Usability is correlated. uppo e interrelationships among System's H 11 : User's Flexible Behavior and System's Flexibility is Partially Flexibility, Users' Flexibility and correlated. Supported Perceived Usability of existing Online H 12 : Perceived Usability of OARS is not affected by Airline Reservation Systems? users' Flexible Traveling Behavior after adjusting for the
RQ6: How do service quality attributes of airlines and external variables jointly predict flexible behavior of travelers?
RQ7: How does user Perceived Usability with the ex1stmg and the proposed system differs?
effect of the covariate, System's Flexibility. H 13 : Flexible behavior of travelers cannot be predicted by service quality attributes and external variables.
H 14 : lJser Perceived Usability with existing and proposed systems is different across the three groups.
RQ8: Is there a multivariate main H 15: There are differences among effectiveness, effect of user's Flexible Traveling efficiency and satisfaction caused by the users' Flexible
Supported
Supported
Supported
Supported Behavior (High, Medi urn and Low) on -o=:To-ra"-v--'e=-=l:c.in-:':g~B.::....:.e=h-av_i-'o-r_. ---:::::--:-------,,-----:---::-----c,------:---:-------effectiveness, efficiency and H 16: Effectiveness, efficiency and satisfaction in the satisfaction of the proposed system? proposed FOARS is highest for users with highest Supported
flexible behavior.
5.1 System's Flexibility
In systems engineering, non-functional characteristics are very important [130].
Normally, Functional Requirements (FRs) and Non-Functional Requirements (NFRs)
are elicited separately but later they are merged together to assess the satisfaction
level ofFRs over NFRs [130]. Sometimes, NFRs are so much dependent on FRs that
their enhancement is not possible. This is especially true in the case of Transaction
Processing Systems where flexibility is sometimes reduced strongly to avoid non
standard operations. Nevertheless, the results of the study suggest that there is a
significant relationship between NFRs and FRs and that NFRs (including flexibility)
have a strong relationship to the usability of the reservation systems.
However, the statistical results indicate that there is a poor correlation between
FRs and the usage of Online Airline Reservation Systems as shown in Chapter 4,
Section 4.1.3. The rationale behind these could be that NFRs are always assessed
against the availability of FRs. However, users are bound to assess the flexibility of
existing online reservation systems on the basis of the FR that they are realistically
exposed to. Therefore, when online users are inquired to respond to questions such as
what additional FR features they like to see in online reservation systems, they may
not conveniently know "what to ask for?" unless they are exposed to such facilities.
On the other hand, NFRs such as the flexibility of the offered features are easier to
assess. This indicates why customers wish to have a more flexible system rather than
having more features.
Furthermore, the data shows that flexibility and other NFR are placed highest in
terms of their Cronbach's a score as shown in Chapter 4, Section 4.1.1, which
otherwise is low, if the data has multidimensional structures. This indicates that online
users perceive the elements of flexibility and NFRs in online reservation systems at
almost equal levels, whereas they perceive the elements of FRs less, these being the
third-ranked item in terms of its Cronbach's a score.
The finding suggested that, first, the incorporation of flexibility has a significant
impact on the usability of online reservation systems as shown in Chapter 4, Section
4.1.4 and, second, FRs and the usability of online systems are perceived to have a
147
meager influence on each other as discussed above. In contrast to the user's
perception, however, one should not overlook. that flexibility is mirrored in functional
requirements as well: Typically, one needs to--so to speak-add more buttons to make
a system more flexible.
5.2 Users' Flexibility in terms of Compromising on SQAs of OARS
One of the major research questions of this study was to investigate Users' Flexibility
towards Online Airline Reservation Systems. As mentioned earlier in Chapter 2,
User's Flexibility is nothing but users' ability to rapidly change from one course of
action to another and it is reflected in users' decision making behavior [53]. For
understanding flexible human behavior, decision making is an important reflection of
users thought process. In the context of User's Flexibility, different users may have
different needs, interests and wishes to be served and system's effectiveness,
efficiency and satisfaction may vary from one user to another based on their usability
perception. For some users a system may be very effective but this may not be true for
all.
The results of this study as shown in Chapter 4, Section 4.2.1 demonstrated that
apparently there was enormous difference between the satisfaction levels of the three
respondents groups, which may not be the actual case. Therefore as a follow-up and to
backup this, the same data was interpreted using Error Bars with 95% Confidence
Intervals (CI). CI of the groups is closely related to the results of the analysis of
variance for these groups. The confidence interval for each graph showed a linear
pattern of the sample distribution which otherwise appeared to be showing huge
variations in the simple means plot as shown in Chapter 4. Section 4.2.1.2. In Error
bars it was examined if mean of one group was included in the confidence interval of
the other two groups. If yes, then it was interpreted that there was no difference
among the groups. It is not relevant whether the error bars 'overlap' but whether the
mean of one group 'overlaps' with the error bars of the other [185]. The confidence
intervals can overlap by as much as 25 percent of their total length and still show a
significant difference between the means for each group. Any more overlap and the
results will not be significant [ 186].
148
T-Test (User's Flexible Behavior as outcome variable) showed signiticant results,
for H6 (t =36.760, df= 169, p = 0.000 < 0.001) and for !11 (t =35.509, df= 169, p =
0.000 < 0.00 I). This indicated the three groups were significantly different in terms of
their rated satisfaction in SB Ts and OT As.
For, H6 and H1 the difference among the mean satisfaction was found to be
statistically significant in ANOV A as shown in Chapter 4, Section 4.2.2.1 and 4.2.4.1
respectively, but that does not indicate which means are actually making the
difference. In other words, three groups of respondents are different, but how they are
different was yet to be determined. Moreover .. in error bar, the size of confidence
intervals for the three groups also differed from one another. When F -test with a
factor that consists of three or more means and additional exploration of the
differences among means is needed to provide specific information on which means
are significantly different from each other, Post hoc tests are performed. Therefore,
ANOV A results of H6 and H1 were further investigated by performing Post-hoc
Scheffe's tests. Scheffe's procedure is the most popular of the post hoc procedures,
the most flexible, and the most conservative [187]. Scheffe's procedure corrects alpha
for all pair-wise or simple comparisons of means, but also for all complex
comparisons of means as well.
Post-hoc Scheffe tests as shown in Table 4.5 and Table 4.8 as shown in Chapter 4,
Section 4.2.2.2 and 4.2.4.2, respectively demonstrated that in both cases (i.e.
satisfaction level with existing SETs and satisfaction level with existing OTAs) there
is a significant difference between the pair of means of the respondents who reported
their flexible attitude as "Cannot compromise" on service quality attributes of the
airline with those who "Can compromise". The same group of respondents also
differed significantly from the group of respondents who reported their flexible
attitude as "May compromise" on service quality attributes.
Furthermore, the effect size show that 19% of the total variance in satisfaction
with existing self-booking tools and 7.5% of the total variance in satisfaction with
existing online travel agencies of the airline is accounted by the flexible attitude of the
users in terms of compromising on service quality attributes of the airline. This is
suggestive of the fact that there is some meaningful difference among the groups and
149
hence the level of satisfaction with existing SBTs and OTAs arc considered different
for respondents with different attitudes towards Users' Flexibility in compromising on
service quality attributes of the airline.
5.3 Integration Assessment of OT As Features into SBTs
Another important research question was related to the integration assessment of
Online Travel Agencies (OTAs) features into Self-booking Tools (SBTs). The
findings of this study as shown in Chapter 4, Section 4.3.1.1 showed that integration
of OT A features would make SBTs more flexible Online Airline Reservation
Systems. This was in accordance with PhoCusWright Report 2009 [9] that reported
the overall share of online travel agencies is improving day-by-day, and was recorded
at 13% in 2008 in US travel market alone and projected to touch 16% in 2011 as
mentioned earlier in Chapter 2. This was further justified !rom Yahoo Travels' claim
which says that 76% of all online travel purchases occur as a result of search function.
The table presented in Chapter 2 (Section 2.1.2.2, Table 2.2) showed innovative
attributes and function that have contributed immensely towards the popular
acceptance of OTAs over Airlines' SBTs and are also widely common among travel
companies in recent years [I], [16].
The findings of this study showed that respondents who considered integration of
OT A features would make SBTs more flexible and also who stated otherwise,
reported absolute lower level of satisfaction with hotel search facility as a
recommended solution for SBTs as shown in Chapter 4, Section 4.3.1.2. The finding
is very much self-explanatory because Hotel Search facility is a popular OT A
integrated feature which displays information of multiple hotels, and promotional
packages available in deals with specific airlines. However. in the context of SBTs,
this feature may not be feasible for integration, since SBTs are self-booking tools
offered by an airline, where multiple carrier and hotel reservation sources are not
incorporated with the reservation planning.
When we look at the other two OTA features, i.e. Matrix Display and Alternate
Airport Search the pattern is quite different and this is the part of interaction effect
!50
that the direction of difference is not in the same direction. Respondents who
considered integration of OTA features into SBTs will not necessarily make them
flexible, opted for Matrix Display and Alternate Airport Search as their recommended
solution for the SBTs. The Matrix display ancl hotel search features are unique in
sense of incorporating multiple sources reservation information. However from
integrating into SBTs perspective, they are not feasible because they require merger
of multiple information resources, which might not be an acceptable standard for an
airline due to its privacy policy and other regulations. On the contrary, alternate
airport search is related to providing additional information as well as the extent to
which a traveler is willing to be flexible in identifying his/her destination sources.
This feature seems practical and has implications for integration into SBTs.
Finally, unlike other OTA innovations, the opaque fare mechanism depends on
hidden characteristics of the traveling plan, thus leveraging upon traveling behavior of
leisure travelers, who are always up for grabs and less sensitive to traveling plans.
This study showed that it is highly recommended by the experts in the airline industry
for integration into SBTs. Researchers point out [117] that opaque products are
flexible in characteristics; therefore, a seller is m a unique position to offer
horizontally differentiated products to customers upon purchase due to the flexibility
of assigning pre-determined products to the customer. Opaque mechanics gained
popularity due to its very unique price discrimination mechanism [ 118] which could
generate incremental revenue for the airline by deliberating upon price sensitive
consumers [119]. In the very short time, opaque selling has attained the status of a
competitive lever for the airline, signifying that an airline could suffer revenue loss to
its competitors by not opting to offer opaque offers [120].
Airlines secure incremental revenue by way of disposing off their distressed
inventory through last-minute sale discounts. The last minute sale discounts are
offered at heavily discounted rates. This mechanism has nothing much of 'opaque' or
'hidden' in it. However, this type of direct selling at the last-minute is considered
very risky for the airline, since potential travelers prefer to wait for last-minute sales
and not purchase in anticipation of heavy discounts [40]. Such a condition may put an
airline in a very risky position with potential possibility of revenue loss. That is why
!51
this practice is substantially criticized by analysts and researchers, who refer to it as a
vivacious cycle of price degradation that can eventually destroy the airlines (11].
Although opaque fare, matrix display, alternate airport search and hotel search facility
are all part of OTAs' innovation, this survey concludes that adoption of opaque fare
concept can make SBTs more t1exible Online Airline Reservation Systems.
5.4 Users' Perception on Factors Influencing Upon Flexible Traveling Behavior
Investigation of Users' Perception on factors influencing upon Users Flexible
Traveling Behavior was also an important research questions. Results of this study as
shown in Chapter 4, Section 4.4.1 indicated that users traveling behavior is molded by
a number of important personality relevant determinants, both internal and external in
characteristics. The findings was supported by other studies mentioning that
flexibility is undeterminable without accounting for traveler's flexible traveling
behavior in the case of Online Airline Reservation Systems [62].
The responses received through online discussion forums as shown in Chapter 4,
Section 4.4.1 identified some very basic key elements that instigate flexibility in a
travelers' traveling behavior, especially if he or she is traveling on a non-rigid
schedule. For instance, a promotional package offered by a particular airline makes it
a lucrative option for travelers who are ready to travel at any alternatively allocated
itinerary. Similarly, if an airline does not offer services and comfort to its customers,
they are likely to become inflexible for opting to fly with it, even if it offers lower
fares. Likewise, travelers flying purpose and flying frequency determines the extent to
which he/she is ready to become a flexible traveler. External influences of travelers'
family and friends also play an important role in shaping up their flexible traveling
behavior. As for the factors that influence upon perception of a flexible online
reservation system, our analysis showed that a reservation system is perceived as
being flexible if it offers easiness to users in terms of making reservations with
minimal skills required, is not complicated in terms of operation and offers multiple
features and options to users.
152
The responses received through in-depth interviews as shown in Chapter 4,
Section 4.4.1 instigated t1exibility in a travelers' behavior, factors such as airlines
repute and travelers status consciousness contribute towards their t1exible traveling
behavior. Interestingly, customer loyalty, traveling mileage and patriotism also came
into the limelight with our in-depth interviews with travelers. This indicates, if a
traveler has a customer loyalty or earned traveling mileage from a particular airline it
will int1uence upon his/her t1exible traveling behavior, as well as, how patriotically
associated the traveler feels with his/her country's airline. Finally, travelers who wish
to live an eco-friendly life supported the idea of the green environment and they were
of the view that if an airline supports eco-friendly policies, with lesser carbon
emissions and lesser fuel consumption, it will int1uence upon their t1exible traveling
behavior as well as t1ying priorities. Likewise, it was revealed that travelers t1exible
traveling is also dependent upon their interest taking in recreational and leisure
activities. Respondents who indicated that they were frequent t1yers attributed t1exible
traveling behavior directly to their interest in recreational and leisure activities. Some
travelers also highlighted that t1exible traveling behavior of a single person or an
individual is different from the one who is traveling with his/her family. As families
with children would prefer to travel mostly during the school breaks or holiday
sessions, therefore, their t1exible traveling behavior will be accordingly int1uenced by
this factor. Travelers during discussion also pointed out the high int1uence of
travelers' occupation on his t1exible traveling behavior, as travelers who are
associated or are in a job that gives them high work load, lesser breaks and requires
more on-job working and lesser mobility, would be less t1exible in their traveling
behavior. As for the factors that int1uence upon perception of a t1exible online
reservation system, our analysis showed that factors such as simplicity and easiness
were again important determinants. From our in-depth interviews, it was found out
that a t1exible reservation system is considered to the one that can provide alternative
dates for t1ying to the traveler and at the same time offers self-adjustment
functionality, such as cancellation and changes, in the itinerary of the traveler. Some
travelers, who were non-native English speakers, pointed that a system would be
t1exible to them if it supports multi-languages.
153
Finally, focus group participants had better and deeper understanding of travelers'
flexible behavior as well as that of reservation system. In-depth interview results as
shown in Chapter 4, Section 4.4.1 with focus groups on factors influencing users'
flexible traveling behavior were very much the same as the one discussed earlier.
However, additional factors that came into the scene included discounted airfares
versus normal airfares and time frame, peak versus off travel season, chosen to fly
also substantially influences upon flexible traveling behavior. Focus group provided a
more critical feedback on factors influencing upon perceived flexibility of an online
airline reservation system and we were able to extract a number of important
additional factors in our analysis.
The data collected from three different sources depicted a similarity pattern,
especially in case of data collected from online travel forums and in-depth interviews.
Data analysis provided a more detailed perspective of travelers' flexible behavior by
incorporating socio-economic factors and societal influences. After giving much
thought process to results as shown in Chapter 4, Section 4.4.1 following 6 themes
emerged under factors influencing upon Flexible Traveling Behavior and 3 themes
emerged under factors influencing upon perceived flexibility of a reservation system.
Theme I: Travelers' Flexible Behavior is molded by their Traveling
Consciousness -Consciousness is a subjective experience of awareness. In context of
flexible traveling behavior, our study showed that travelers have a certain form
consciOusness or awareness about their traveling behavior. This resulted in the
emergence of first theme.
"I look.for great deals through promotional schemes" (ID P 2)
"Why I wouldn't travel with an airline that provides heavy discounts? I look for
discounts all the time" (ID P6)
"I make cognizant traveling decisions. Do all background workfirst and then make a
decision to purchase a ticket or not" (OTF PI 1)
"Airlines hefty discounts through their travel mileage schemes are very attractive and
can make me to travel any time during the year" (FG P7)
154
Theme 2: Travelers' Flexible Behavior is molded by their Belief that they
have the Required Digital Skills - Study showed that having the digital skills to
search for the best traveling deals, can influence upon once flexible traveling
behavior. Digital skills reflect upon travelers' self-efficacy and have resulted in
evolution of our theme 2.
"I don't care traveling with any airline, as I canfind best traveling online deals with
much ease" (OTF P8)
"I don't waste time on googling what I am looking for, I know where to exactly go
when I need to purchase an online ticket" (ID P 26)
"It takes patience and perseverance to find out best traveling deals" (OTF P3)
Theme 3: Travelers' Flexible Behavior is molded by their Self-belief as
Flexible Travelers - This theme was categorized on the basis of how respondents
reflected upon their self-identity as travelers. Two sub-themes emerged.
Sub theme 1: Traveling Motivation (traveling purpose, flying frequency, interest
in recreation).
Motivation to travel is an important factor to influence upon travelers flexible
behavior. For instance;
"It very much depends upon my traveling purpose and motivation" (ID P6)
"I love to travel, this is my biggest motivation" (OTF PI)
"As a sales man, if one has to travel around the year, his traveling motivation will be
lower" (FG P 15)
Sub theme 2: Socio-economic Position
Socio-economic standing of the traveler also influence upon flexible traveling
behavior.
"depends on the job you are in" (OTF P4)
!55
" ... with limitedfinancial resources, one can only ion!{ to take leisure trips" (ID P2)
"I am more flexible in travelinf{ when I am paying/or my ticket and can also
compromise on comfort. This is not the case when my office is paying/(1r me" (FG
P9)
Theme 4: Travelers' Flexible Behavior is molded by Societal Influences -
This theme emerged due to the social influences that influence upon a traveler's
flexible traveling behavior. Social influences occur when a traveler's flexible
behavior is affected by other people.
"We travel with our two sons. So our preference is to travel during their school
holidays" (IDPI6)
"We lookfor cheap deals as we have a large family to support and travel together"
(OTA PI7)
"My father reserves ticket for me when I travel or I call my friend in a travel agency.
They decide for me. Simple 1" (ID P I4)
Theme 5: Travelers' Flexible Behavior is molded by how they attribute A
Cause to their Traveling Behavior - This theme emerged out a certain group of
respondents who attributed a cause to their traveling behavior.
"We prefer Malaysian airlines over other airlines. It is our pride" (OTF P I2)
"I prefer airlines with high repute, since they significantly enhance my traveling
experience and pleasure" (ID P9)
"I am a supporter of eco-friendly living. I would he flexible to travel with an airline
that supports this great cause" (ID P I4)
"Airlines repute fascinates me a lot" (FG 8)
Theme 6: Travelers' Flexible Behavior is molded by their Prior Traveling
Experiences - Surely, traveler's prior traveling experiences influence upon their
!56
future traveling decisions. This theme has much to do with traveling comfort that one
expenences.
"I prefer airlines with better leg space area and comfortable seats" (ID P 30)
"Sometimes customers compromise on jet ways availability" (FG P 18)
"Staff attitude is a traveling motivator" (OTF P 11)
"We prefer airlines that serve Muslim halalfood" (OTF P21)
Theme 7: Systems Perceived Flexibility is influenced by its Perceived
Usability - A System will never be questioned for flexibility, if it satisfies users'
expectations. Therefore, systems perceived flexibility is always dependent on the
usability of a system.
"Searching/or good packages using existing OTAs and SETs is quite simple and
easy" (ID P8)
"It requires the fewest steps to accomplish what users want to do" (FG 15)
"Users would prefer to use our system because they can make changes in their
itinerary very easily" (FG 9)
Theme 8: Systems Perceived Flexibility is influenced by End User Support -
Our study showed that post -sale features are very essential in determining the
perceived flexibility of a system. This theme has much to do with the user support.
"Every online reservation ;,ystem has the capability to reserve a ticket ... ... for me a
flexible system is the one which provides after sale service features" (/D 16)
"A flexible reservation system should support user prompting and guidelines
throughout the reservation process" (FG P20)
Theme 9: Systems Perceived Flexibility is influenced by Comparison of
Features on the Actual Level of Effect regarding to Complete the Reservation
Process - This theme emerged due to the comparison features that influence upon
157
systems perceived flexibility. Provision of multiple features in search of a better flight
fare on different dates enhances the systems perceived flexibility.
"Me as a user would like to compare flight Jares on dijji!rent dates" (ID P22)
"It may (system) indicate about multiple destination options to users to choose from
and accordingly point out fare changes" (FG P 17)
"A system providing multiple searching options, like by dates, fares, destinations"
(ID Pl4)
"If required, changes in itinerary can be made or at least recommended by the
customer to the system" (ID P21)
Our analysis showed that an Online Airline Reservation System is perceived as
flexible if it provides fare quotes, post-sale features and generates receipt. Likewise, if
the reservation system can sort airfares on different dates and destinations, it will be
perceived as flexible by users. Moreover, if the user is prompted throughout
reservation process for his/her guidance, is timely notified on low airfares though
flexible and alternative flying dates, it will influence upon user's perception of a
flexible reservation system. Also by incorporating multiple traveling components in a
single search, such as hotel search facility, the perceived flexibility of the reservation
system will also be influenced.
The findings of the study showed that users traveling behavior is molded by a
number of important personality relevant determinants, both internal and external in
characteristics. While, traveling consciousness, self-efficacy in digital skills and self
belief as flexible travelers are internal personality relevant determinants influencing
directly upon travelers flexible behavior, societal influences, attribution and prior
experiences are external personality relevant determinants that indirectly influence
upon travelers flexible behavior. Moreover, external determinant may not necessarily
always have the same influence every time, depending upon the situation a traveler is
m.
158
5.5 Classification of Users on the Basis of Their Flexible Traveling Behavior
The most important research question of this study was the categorization of users on
the basis of their flexible traveling behavior. Results as shown in Chapter 4, Section
4.5.1 revealed that users can be classified as High, Medium and Low flexible on the
basis of Users' Flexibility score. The study results was supported by other researchers
[ 61] since consumer behavior researchers have frequently employed schema theory as
the theoretical underpinning of their investigations for classification of consumers as
with a likelihood of high purchase power, medium purchase power and low purchase
power.
The results further revealed that Users' Flexible Behavior is positively associated
with Perceived Usability of Flexible Online Airline Reservation Systems as shown in
Chapter 4, Section 4.5.2. A traveler is categorized as highly flexible, if he/she
possesses a flexible nature of traveling [188]. This indicates that he/she has a non
rigid schedule for traveling. Flexible travelers can travel on alternatively allocated
itineraries and their traveling satisfaction comes mainly from heavy discounted fares
provided by any airline. On the contrary, a traveler is categorized as low flexible, if
he/she possesses an inflexible nature oftraveling. This indicates that he/she has a rigid
schedule for traveling. Low flexible travelers only prefer to travel as per their
indicated or preferred itineraries and their traveling satisfaction comes mainly from
meeting their flying deadlines that are mostly based on reaching their destination as
per their preferred itineraries. This indicates that as Users' Flexible Behavior
improves (from Low to Medium I from Medium to High). their Perceived Usability of
Flexible Online Airline Reservation Systems tends to also improve.
To understand causality of interrelationships between Users' Flexible Behavior,
System's Flexibility, System's Adaptability, System's Adaptivity and System's
Personalization, conclusions are drawn about variability by squaring the correlation
coefficients. By squaring the correlation coefficient, a measure of how much of the
variability in one variable is explained by the other can be derived.
Although, System's Flexibility share a non-significant correlation coefficient with
Users' Flexible Behavior, two of its sub-variables, System's Adaptability and
159
System's Personalization happen to be sharing significant correlations. The value ofr2
for System's Adaptability is (0.372)2 = 0.138. and for System's Personalization is
(0.314) 2 = 0.098. This explains how much of the variability in Users' Flexible
Behavior is accounted for by two of the sub-measuring variables of System's
Flexibility. In percentage terms, System's Adaptability accounts for approximately
14%, while System's Personalization accounts for approximately 10% of the
variability in Users' Flexible Behavior and together the two variables account for
approximately 24% variability.
5.6 Role of SQAs and External Variables in Flexible Behavior of Travelers
The Pearson Correlation Analysis as shown in Chapter 4, Section 4.6.1.2
demonstrated that all correlations are statistically significant. This may help explain
the inconclusive results found by other researchers [89] to determine the contextual
interpretation of any event by contextual factors that reinforce viewers' schemas,
formulate characteristics of the surrounding environment and ensure effective
collaboration between the two.
The results on scatter plot examining service quality attributes of the airline
against users' flexible personality as shown in Chapter 4, Section 4.6.1.1 are normally
distributed along with linear regression line and have no obvious outliers. 5
represented being least important in terms of compromising on service quality
attributes of the airline, while 5 also represented being highly flexible in nature.
Therefore, with every point increase in compromising on service quality attributes of
the airline, users' flexible behavior tends to increase linearly. The R square for the
best fit line is 62.7% indicating a very strong positive relationship between the two
variables. This further means that about 62.7% of the variability in Users' Flexibility
is accounted for by the service quality attributes they are ready to compromise on or
to forgo. Similarly, the results on scatter plot examining external variables against
users' flexible personality as shown in Chapter 4, Section 4.6.1.1 are normally
distributed along with linear regression line and have no obvious outliers. 5
represented strongly agreeing to the influence of a particular external variable, while 5
also represented being highly flexible in nature. Therefore. with every point increase
160
in strongly agreeing to the external variable influences, users' flexible behavior tends
to increase linearly. The R square for the best fit line is 63%, which indicates a very
strong positive relationship between the two variables. This further means that about
63% of the variability in Users' Flexibility is accounted for by the external variables
and their influences upon their traveling behavior.
The value of the R' as shown in Chapter 4, Section 4.6.1 is a measure of how
much of the variability in the outcome is accounted for by the predictors. Its value is
0.669, which means that the two variables approximately account for 67% of the
variation in predicting users' flexible behavior. The adjusted R' gives some idea of
how well this model generalizes and the closer its value is to R', the better it is
considered for model fitness. In this case, difference for the model is reasonable
(0.669- 0.667 = 0.002 or 0.2%). This shrink.ag<: means that if the model was derived
from the population rather than sample, it would account for approximately 0.2% less
negligible variance in the outcome.
In Multiple Regression, the model takes a form of an equation, which has a
coefficient (b values) for each predictor variable. The first part of the table as shown
in Chapter 4, Section 4.6.1.5 estimates b values which indicate individual contribution
of each predictor to the model. The b value signifies relationship of each predicting
independent variable in Users' Flexibility. A positive coefficient indicates a positive
relationship, while a negative coefficient indicates a negative relationship. The b
value also tells us to what degree each predicting independent variable affects Users'
Flexibility, if the effects of all other predictors are held constant. The highest positive
standardized beta value is for the external variables, namely "Persuasion" (0.264)
closely followed by "Engagement" (0.166). This indicates these two independent
external variables have a strong impact on determining users' flexible behavior, with
variable former having a slightly higher impact. This is followed by a Service Quality
Attribute of the airline, "Last minute discounte:d airfares" (.128), which also tends to
have a hefty influence upon users' flexible traveling behavior. External variable
"Identity" also makes it into the top contributing factors with standardized beta
coefficient (0.125). Finally, Service Quality Attributes of the airline, "Immediate
Confirmation of Itinerary on purchase" and "Confirmation of Origin and Destination
161
Airport" were found to have significant impact on determining Users' Flexibility with
beta coefficients of (.119) and (.113 ), respectively.
After examining direct relationship of each predicting variable with Users'
Flexibility, the b value of each predicting variable in association with its !-test statistic
and significance level was interpreted. If the t-test associated with the b value is
significant at either p<O.l, 0.5 or 0.001 level, then that predictor is making a
significant contribution to the model. For this model, external variables Persuasion (t
(249) = 1.335; p = .018 < .05), Engagement (t (249) = 2.548; p = .011 < .05), Identity
(t (249) = 1.638; p = .037 < .05) and Service Quality Attributes of the airline, last
minute discounted airfares (t (249) = 1.829; p = 0.43 < .05), confirmation of origin
and destination airport (t (249) = 1.572; p = .017 < .05) and immediate confirmation
of itinerary on purchase (t (249) = 1.542; p = .003 < .05) are significant predictors of
Users' Flexibility.
5.7 User Perceived Usability with Existing and Proposed Systems
Researchers evaluate Perceived Usability of Online Airline Reservation Systems by
doing content analysis on the basis of effectiveness, efficiency and satisfaction [ 10 !],
[I 04]. It is also important to investigate web users' characteristics, preferences,
expectations [I 08] and their behavior from online systems, and to compare with the
Perceived Usability of the systems. The findings of the study as shown in Chapter 4,
Section 4.7.1 demonstrate significant differences on effectiveness, efficiency and
satisfaction among the classified users on the basis of their flexible traveling behavior.
The study revealed significant results F (2, 244) = 22.588, p < .001 for the
respondents who indicated that there is a significant main effect of the users'
classifications (High, Medium, and Low) on perceiving effectiveness of the proposed
and existing systems. This effect indicates that overall, when the influence of existing
and proposed system is ignored, the users' classifications on the basis of their flexible
traveling behavior (High, Medium, and Low) int1uenced upon how they perceive
effectiveness of a system. When the effect of existing and proposed systems was
examined on perceived effectiveness, there was a significant main effect, F (I, 244) =
162
26.627, p < .001. This effect means that with respect to ignoring the classification of
users on the basis of their flexible traveling behavior, the proposed and existing
systems influenced upon users' perceived effectiveness. Furthermore, F-test results
further reveal a significant interaction between the effect of existing and proposed
online reservation systems and the user classifications on perceived effectiveness, F
(2, 244) = 15.612, p < .001 as shown in Chapter 4, Section 4.7.1.2.
The second F-test for respondents, who indicated that there is a significant effect
of user's classifications on the basis of flexible traveling behavior, there was a
significant main effect, F (2, 244) = 6.065, p < .05. This effect indicates that overall,
when the influence of existing and proposed system is ignored, the users'
classifications on the basis of their flexible traveling behavior (High, Medium, and
Low) influenced upon how they perceive efficiency of a system. When the effect of
existing and proposed systems is examined on perceived efficiency, there was not a
significant main effect, F (1, 244) = 0.545, p = .461. This effect means that if we
ignore the classification of users on the basis of their flexible traveling behavior, the
proposed and existing systems does not influenced upon users' perceived efficiency.
Furthermore, F-test results further reveal a significant interaction between the effect
of existing and proposed online reservation systems and the user classifications on
perceived efficiency, F (2, 244) = 4.878, p < .05 as shown in Chapter 4, Section
4.7.1.6.
This was followed by the third F -test examining the respondents, who indicated
there is a significant effect of user's classifications on the basis of flexible traveling
behavior, there was a significant main effect, F (2, 244) = 3.476, p < .001. This effect
indicates that overall, when the influence of existing and proposed system is ignored,
the users' classifications on the basis of th(:ir flexible traveling behavior (High,
Medium, and Low) influenced upon how they perceive satisfaction of a system. When
the effect of existing and proposed systems is examined on perceived satisfaction,
there was a significant main effect, F (1, 244) = 8.728, p < .05. Furthermore, F-test
results further reveal a significant interaction between the effect of existing and
proposed online reservation systems and the user classifications on perceived
satisfaction, F (2, 244) = 7.372, p < .05 as shown in Chapter 4, Section 4. 7.1.10.
163
5.8 Multivariate Main Effect of User's Flexible Traveling Behavior on Proposed
System's Effectiveness, Efficiency and Satisfaction
One of the major research questions of this study was to investigate the effects of
user's flexible traveling behavior on the proposed system's effectiveness, efficiency
and satisfaction. The findings of the results as shown in Chapter 4, Section 4.8.1
suggested that users with highly flexible behavior differed significantly from users
with medium and low flexible behavior in terms of rating effectiveness, efficiency
and satisfaction of the proposed system. This may help explain the inconclusive
results found by other researchers [101] saying that Usability of a website cannot be
improved without considering consumer intend and behavior.
Researchers [34]-[36] argued that clear understanding of consumer intent and
behavior in the case of online airline ticket shopping and elsewhere cannot be
achieved without considering the factors that affect purchase decisions. Results as
shown in Chapter 4, Section 4.8.1.2 prove that the effectiveness of the proposed
system differed significantly from users with highly flexible behavior with medium
and least flexible behavior. Similarly, in the case of efficiency of the proposed system,
users with highly flexible behavior differed significantly from users with least flexible
behavior. Furthermore, satisfaction of the proposed system also differed significantly
from users with highly flexible behavior with medium and least flexible behavior.
5.9 Recommendations for Flexible Online Airline Reservation Systems
Research findings show that system's perceived flexibility is reflected in its Perceived
Usability. Perceived Usability is a combination of system's effectiveness, efficiency
and satisfaction. End user support or user prompting is considered to be a supporting
characteristic that substantially augments efficiency and effectiveness of the system,
while multiple options directly influence upon user satisfaction. If a system provides
ranges of dates as flying and source destination options at different fares, flexibility of
the system is enhanced in its Perceived Usability in the eyes of the users. This is
because if a user chooses a flying option 'A' from a given one or two options, he has
not made a flexible decision. But if he chooses the same flying option 'A' from a
164
variety of given flying options, he is likely to enjoy extra satisfaction that he will get
from the flexibility of the system and also in his purchase making decision. Likewise,
if a system offers multiple flying options, they will also influence upon users'
decision and make them change their mind to opt to fly from option 'B' instead of
'A'. This will again have positive influence upon user's satisfaction from the system
Perceived Usability.
This thesis suggests a new approach of reservations while introducing high,
medium and low flexible travelers. Traditionally, flights schedule is made available
for the travelers after finalizing the resources such as, planes, airports and dates.
Travelers make reservations on their favorite dates using the offered flight schedule as
shown in the Existing Approach of Figure 5.1. Current approach of airline reservation
systems is adequate in terms of making reservations. However, traditional approach
does not provide an opportunity to travelers of being flexible and also it does not
provide any support to improve load factors in the flights, and therefore, airlines
either cancel the flight [189] at 11th hour if they receive less booking in any particular
day or they send flight with minimal profit margin.
In this study travelers are categorized on the basis of their flexible traveling
behavior as shown in the Proposed Approach of Figure 5.1. In the proposed
framework, least flexible customers may choose confirmed schedule offered by the
airline with normal fare. Least flexible customers will get their confirmed tickets with
seats allocated and itinerary finalized at the time of booking. Since flexible travelers
are flexible enough and have not indicated their preferred dates to fly, therefore, they
can be requested to provide their preferred time-frames (e.g. Departure: any flight
between lOth June to 20th June from Kuala Lumpur to Frankfurt, Arrival: any flight
between lO'h July to 201h July from Frankfurt to Kuala Lumpur) in which they would
like to travel. Tickets to highly flexible customers will be issued at booking time but
the seats and flight day will be notified only a week before their actual departure day.
Thus, it becomes the responsibility of the arbitrator systems to manage and finalize
their traveling itineraries as per their indicated time-frames. Once done, the flexible
customers can be allocated seats a few days before their actual departure by the
airline.
165
0\ 0\
* Yluih1•
Leg•C)- Systems
Traditioaal .'\pproacla Propos-.d .'\pproaelt
Figure 5.1: Framework for Flexible Online Airline Reservation Systems
* lafluibl ..
5.9.1 Graphical User Interface for FOARS
The recommendations for the design of Flexible Online Airline Reservation helped in
developing the Graphical User Interfaces as shown in the following sections.
5.9.1.1 Home Page ofFOARS
Figure 5.2 shows interface for the Home Page of Flexible Online Airline Reservation
System. The description of Flexible and Inflexible passengers as shown in Figure 5.2
will help users to understand the difference between the two terminologies used for
booking purpose. Passengers who are flexible with regards to flying dates will be
treated as Flexible Passengers and those who wish to travel on specified dates will be
treated as Inflexible Passengers.
On 2tl- Aug.st :!(110, "'exT .. tWIJY!i WO$ 05labiowd w<h to hope:>' moCI'; tne IOW'Iabat' l'lGr< flevob e.-- 1"010 cl;-jOCtrve Gf l'texT'-""'0)" IS ro pro..O. :\\'0 N•Y ll<rtf.t:J, fWa. d-up fli!jht for<S f01 'lexltle :r .. ,eiJ.'S, >rod~ • ..-.ding ro;t-u ovtll ~ copaa:y l.lbltzotio• v:tuch "' :.n1 coold rtduco me number ol rlg~ls 1!\o' oc:uolly take-olf per ... a< Subsequt11tly ~ng tl'o proT. "'•r;•n for the o;tit>e •r6JOtrf. F~TA""IJYS wo.Jid •reet der>o'XS ard "-~Y n-ore n.o..tlv by lt.lllzlng
,...,.... .. \'With rraxw1...,. upxtv "1tl1 the ~olp ef passer.gers daS>lb!lon as '"'""'"lr
lnlk>nble possenter I Flexible .,.uenger
I P.....-gon who pre~":::::. tD lly "'.,."',...~_,_,a->~" ore fl<_'llble ~':n ~ords tD flyonv aates.
1 ""'""'' olftrtd •ere (too , o1 me ~~ peke) 1 01=\.nttd rare c•.o-20% disc"""' on tnt normal t.re>
I .\lloca:o- of ..... plane and date v.-,1 boat tnt of I Alloclmor d ..... plane.,.., dW! •• il be Oont r ... days ,...,..!ion. (<-9· 10 oa;-s) before me""""' cjepa"<J'O .. ..
I§!§M
Figure 5.2: Home Page for Flexible Online Airline Reservation System
167
5.9. 1.2 Booking Interface of FOARS
The users of FOARS will be provided two options for Flight Type i.e. Inflexible and
Flexible as shown in Figure 5.3. If users opt for Flexible Flight, they will be requested
to provide the number of days in which they will be flexible as shown in Figure 5.3
?
I ' /~
Online Booking
!From 1 I::L-- -fro: -!Date: 10.11-2010 -[Flight Type : [ '" l~lullle ·-lflexlble oays : I• ~.,.vs dlotounts
!Adult : ,. PJ •· •~ [Children ,--~~l J ••" ~ I Infant : I o - ·-· faa .. rype : _lE"""""' -
-. ...
Figure 5.3: Booking Window for Flexible Online Airline Reservation System
5.9. 1.3 Flight Search Window for Flexible Travelers
Flight search with Flexible Flight Type will let the users to the searched window with
flights available within the flexible dates as shown in Figure 5.4.
II You are • flexible pusenger
fo~scount : 120'141 tfo1111lo pay : [MYR1725150
o Fllghls ••~liable within your flexible days.
You will be alloaoted to one of them.
(You will be lnfonn 10 days before your flight)
~I From rr.;-r Date rn.;;;-r StaltK
[ll ICuola lumpur I [)uba. I t2/ll/10 I ooJO I ....... -fTI KU1IIo Lumpur I Dlbal I 1~/11/10 fli"OOI .... .-f31 Kualltlumpuo ~I !S/liJlO I oo:30 I """""'*
' '
Figure 5.4: Flight Search Window for Flexible Travelers
168
5.9.1.-1 Travelers· Details Window f or Flexible Travelers
Traveler will be requested to fill in the required information for the booking as shown
in Figure 5.5 .
o You are a flexible passenger
Flight
From ~~ Flexibl<! O.Oys I Djscount I Oass
Kuola l'-"'!llr ~~ 6 I 20'Mt I Economy
l"ad Traveller
lrull Name : vr ..
loate of Birth : I · . ... ..• .., . '9S1 •
'Sex : "•• . !Nationality :
!Passport Numbe< :
• • I Mobile T el<>phone :
lEman Address :
Figure 5.5: Travelers' Details Window for Flexible Travelers
5.9.1.5 Payment Window for Flexible Travelers
Users will be requested to make payment with discounted air fare using the possible
options as shown in Figure 5.6.
~> flgf1 RegutrfQ)fnt > Trnyellers Oeto!k > Payment l-tethod
o You are a flexlb~ passenger
flight
From r::=r,;-[ flexible Days I Olscount I Cla!i5
Kt..., l umpur ~~ 6 I 20'Mt I Economy
Payment Details
Ieard Type : I ~.ta.w Co':S •
[a.rd Nurnbe< :
la.nt Holder :
la.rd Holder Telephone :
"""'"' tick this box to K cept the
•=m• •=m• Figure 5.6: Payment Method for Flexible Travelers
169
5.9.1.61mmature Ticket Window for Flexible Travelers
Users will be issued a receipt of payment as shown in Figure 5.7. However. the
allocation of flight out of the given flights wi II be done 10 days before the departure
date.
~ » f!Klht RCO\MTfMC"Dl » lrb·~trtca Peta!l$ > Pllynle1lt Mftbod > Travet~rs P1 tnt Deli:1~15
I Travellrn Of>talfs (flvc:lble)
I Booking I 0 : jAtOOl IP11Stport Num~r : 112345678910
INome: I>< liS GeMn YOhyO ..,_,
I From: IKuolo LUmpur
ITO I 'Cobol 'Clau l)fpe : Economy
'F lexible Days : 1:-.~
OIKOunl.:
ITot.l to P'IY : I><YR1725.60
I PoMibkt D•t~ for your FIIQht
I-:' if"~--~ I Date fTrme I Status
fl'l Kuala lumpur I Oubol I 12/11110 I oo:Jo T A"Y•J.Lib&e
Ill K~lumpw 1 eobo• 1 14111/10 ~I A.va1iable
Ill KUlla Lumpur I Oubll I 15/11/10 roo;o-1 Av.._ttble
Figure 5.7: Immature Ticket for Flexible Travelers
5.9.1. 7 Flight Search Window for Inflexible Travelers
Flight search with Inflexible Flight Type as discussed in Section 5.9.1.2 will let the
users to the searched window with available flights as shown in Figure 5.8.
J
o You ore on lnllexlble pusenger
~rom: uala Lumpu~
o: Dubol
~eqe_u5t Oate 1 Cl'l/11/10
0 night Is not available on your requested elates.
The dosest a-nl~bte d•tes have: been shown Instead
• =m• e:mw Figure 5.8: Flight Search Window for Inflexible Travelers
170
5. 9.1.8 Travelers' Details Window for Inflexible Travelers
Traveler will be requested to fill in the required information for the booking as shown
in Figure 5.9.
o vou ~r"''l!; 3n Inflexible p.-,ssenge.r
I Flight 10 I I FAlOO I
Flight
l..e.>CIT-
!f-ull Name~ I Date ol Olrth I• . .... ..., • "957 • - .
~----~ 1~- , ~~H-~--T-.~~~-~~no-.-. --:r-------------------------1 I Mobile phone:
lEman Address :
Figure 5.9: Travelers' Details Window for Inflexible Travelers
5.9.1.9 Payment Window for Inflexible Travelers
Users will be requested to make payment with actual fare using the possible options
as shown in Figure 5.1 0.
Icard llokler :
!card Holder Telephone :
Please tide this box to accept the
l :ffl!M
Figure 5.10: Payment Method for Inflexible Travelers
171
5.9.1.10Ticket Windowfor inflexible Travelers
Users will be issued a ticket with the allocation of fl ight, seat and other relevant
information as shown in Figure 5.11.
? ,
~ :)o flklbt muarncnt > ~ > lcJYdf1) Qetjif's J> ~ MClbgd ,. T~J Pnnt Odbi!S
I Trav<llen DdaiiS (Jnll<!xillle)
l&ooklno 10 : jA1001
!Seat No : IFAAIO
!Passport Number : luJ<s
!Name : IMtss Ooyang R<>turyo B<ntl A"'M>IJ R.ombll
lnlght to : lrAJOO
!From : l teua~ lumpur
Fo:~e : IOUOOI '11/ 11/10
lnme: loo:JO
!class Type : I economy
!Total to P~v : IMYIUJS7.00 -
J
Figure 5.11: Confirmed Ticket for Inflexible Travelers
5.10 Chapter Summary
Results obtained during the study were discussed in this chapter. Results supported
most of the hypothesis achieving all the three research objectives. In the first place
System's Flexibility and Users' Flexibility were discussed to investigate the user
needs associated with FOARS. This was followed by the discussion on the
classification of users on the basis of their Flexible Traveling Behavior. Later, the
results obtained from a case study to test the validity of the proposed framework for
designing more FOARS was discussed. This study concludes that although System' s
Flexibility and Users' Flexibility are unequivocally independent of another, in terms
of designing flexible Online Airline Reservation Systems, they are part of single
quantum. While User' s Flexibility is related to their purchase making decisions,
Systems Flexibility, although is reflected in its Perceived Usability, can influence
upon user' s traveling behavior by enhancing their satisfaction. Finally, the
recommendations for the designing of a more flexible Online Airline Reservation
Systems were discussed.
172
6.0 Chapter Overview
CHAPTER 6
CONCLUSIONS
This chapter presents the summary of the thesis with major findings. Section 6.1
presents the dissertation summary while summarizing the overall thesis. Section 6.2 is
dedicated for the major research findings and Section 6.3 is devoted for the future
work.
6.1 Dissertation Summary
Behavioral characteristics in terms of making travelers flexible are appealing and an
important area of research. This research provides a comprehensive study focused on
what makes a traveler flexible on the basis of his/her behavioral characteristics, hence,
giving a crucial insight for designing of flexible Online Airline Reservation Systems.
This research therefore, contributes to the small but growing literature of designing
flexible Online Airline Reservation Systems and provides a general framework by
molding upon flexible behavior of travelers.
In this research a framework for flexible Online Airline Reservation Systems has
been proposed, based on integrating opaque characteristics into the SBTs of airlines,
to address the research gap as stated in Chapter l. The framework is based on
categorizing travelers on the basis of their flexible traveling behavior. Travelers with
least flexible behaviour may choose confirmed schedule offered by the airline with
normal fare. Least flexible customers will get their contlrmed tickets with seats
allocated and itinerary finalized at the time of booking. Since flexible travelers are
flexible enough and have not indicated their preferred dates to fly, therefore, they can
be requested to provide their preferred time-frames in which they would like to travel
instead of providing fixed schedule. Tickets to highly flexible customers will be
issued at booking time but the seats and flight day will be notified only a week before
their actual departure day. Thus, it becomes the responsibility of the arbitrator systems
to manage and finalize their traveling itineraries as per their indicated time-frames.
Once done, the flexible customers can be allocated seats a few days before their actual
departure by the airline.
To develop the framework, existing Online Airline Reservation Systems were
examined to assess the flexibility and usability of the systems. The overall research
was divided into three phases and each phase contains one core research objective
which was achieved through quantitative and qualitative techniques to assess
System's Flexibility, Users' Flexibility and Usability of the systems. A redesign
solution for enhanced usability was developed based on HCI guidelines and the
flexibility tactics used in online travel agencies. A new flexible Online Airline
Reservation System design was applied, which led to a proposed interface with the
integration of opaque mechanism. The two interfaces were used in the case study.
Participants were requested to complete the evaluation of the existing and proposed
interfaces.
The benefit of framework given in this research can be used as a baseline for
further studies and to design more flexible Online Airline Reservation Systems.
Moreover, the Users' Flexibility measuring scale established and tested in this
research is also a significant research contribution for future studies to measure
flexibility. This research builds on previous and ongoing work within the disciplines
of Human Computer Interaction by introducing psychometric scales to measure
Users' Flexibility in terms of compromising on service quality attributes of an airline.
Furthermore, this research introduces a new approach of reservations to increase the
passenger load factor by leveraging upon travelers' flexibility. The Information
System (IS) theory given through qualitative reporting is based on empirical findings,
which is novel in its own right and testable in different environments.
174
6.2 Research Findings and Contributions
This study concludes that although System's Flexibility and Users· Flexibility are
unequivocally independent of another. in terms of designing flexible Online Airline
Reservation Systems. they are part of single quantum. While User's Flexibility is
related to their purchase making decisions, Systems Flexibility. is reflected in its
adaptability, adaptivity and personalisation. When considered together, the two can
immensely enhance systems' perceived usefulness, as retlected in its effectiveness,
efficiency and satisfaction.
The overall aim of this research was to investigate the associations between
System's Flexibility, Users' Flexibility and Perceived Usability of Online Airline
Reservation Systems and to propose a framework for flexible Online Airline
Reservation Systems. This research has been conducted to answer the research
questions and achieve the desired research objectives. The objectives were used as
guidelines for the development and analysis of proposed framework. The findings of
this research based on the objectives are summarized below:
6.2.1 Objective 1
To assess user needs (System's Flexibility and Users' Flexibility) associated with
Online A ir/ine Reservation Systems.
Findings- The literature was reviewed that revealed the facts that OTA's success
immensely depends upon hidden or opaque characteristics. It was further revealed that
an opaque characteristic depends on flexible behavior of travelers. Hence, it is
important for the airlines to design their self-booking tools in view of customers'
preferences, expectations and online usage behavior to run successful business.
Therefore, the same concept of opaque mechanism was adopted for integration into
SBTs of airlines and considered a potential research gap. The framework proposed in
this study moulds upon flexible traveling behaviour of users' too and integrated
opaque characteristics in designing of flexible Online Airline Reservation Systems.
175
Since flexibility is referred to its ability to respond to internal or external changes.
Change can be defined as the transition over time which requires change a gent.
Researchers argued that "if the change agent is external to the system, then the change
under consideration is a .flexible-type change" [12]. Therefore, in case of SBTs
incorporated with opaque fares would serve the role of external change agent by way
of providing flexibility in users' decision making. Similarly, "if the change agent is
internal to the system, then the change under consideration is an adaptable-type
change" [12]. Thus the provision of opaque fares into SBTs also serves the role of
internal change agent by way of providing the capability of accepting changed
decisions. If no change agent exists, then the system is rigid (no change can occur).
Since provision of opaque fares could make users flexible and also increases the
adaptability of the system, it is expected that the usability of the system would be
enhanced.
The opaque fare mechanism depends on hidden characteristics of the traveling
plan, thus leveraging upon traveling behavior of leisure travelers, who are always up
for grabs and less sensitive to traveling plans. The findings showed that it is highly
recommended by the experts in the airline industry for integration into SBTs.
6.2.2 Objective 2
To propose a framework for designing more flexible Online Airline Reservation
Systems while classifYing users on the basis of their Flexible Traveling Behavior.
Findings - The research revealed a framework for flexible Online Airline
Reservation System based on the research scope as presented in Chapter 1. The
framework is a general framework that can be applied to different reservation systems
particularly, the Airline Reservation Systems due to their immense profit making
potential and research gap as identified in Chapter 2 and framework methodology in
Chapter 3. The framework is based upon investigating three research variables namely
Users' Flexibility, System's Flexibility and Perceived Usability of Online Airline
Reservation System. User's Flexibility is based upon the extent to which they can
forgo or compromise on service quality attribute of the airline. By doing this users'
176
classification based on their mental models was done for randomized experimentation
of the existing and proposed interfaces. As for System ·s Flexibility the proposed and
existing interfaces used in experimentation differed hy way of leveraging upon
opaque characteristics as offered by OT As. The opaque characteristics in the proposed
t1exible systems will positively enhance perceived usefulness of the systems.
6.2.3 Objective 3
To study the interrelationship between System's Flexibility, Users' Flexibility and
Perceived Usability of Online Airline Reservation Systems and to determine the
Perceived Usability oft he existing and proposed systems.
Findings- The research findings on the interrelationship between Users' Flexible
Behavior and Perceived Usability showed positive relationship. This means that as
one variable increases in value, the second variable also increases in value. This is
called a positive correlation. The significance value indicates that the relationship
between the two variables is genuine, hence both variable share positive relationship
between each other. The overall relationship between User's Flexible Behavior and
System's Flexibility showed non-significant values. However, there were significant
relationships between User's Flexible Behavior and the two components of the
System's Flexibility i.e. System's Adaptability and System's Personalization.
Therefore, both variables i.e. System's Flexibility and Users' Flexibility are partially
associated.
Furthermore, findings proved that System's Flexibility predicts Perceived
Usability of the Online Airline Reservation Systems. What was more interesting is
that when the effect of System's Flexibility was added, the effect of Users' Flexible
Behavior remained significant towards predicting Perceived Usability of t1exible
Online Airline Reservation System.
Moreover, the findings of the study showed significant differences on
effectiveness, efficiency and satisfaction with existing and proposed systems among
the classified users on the basis of their Flexible Traveling Behavior. When the
int1uence of existing and proposed system was examined on perceived effectiveness,
177
efficiency and satisfaction there was a significant main effect with effectiveness and
satisfaction. This effect means that if we ignore the classification of users on the basis
of their Flexible Traveling Behavior, the proposed and existing systems influenced
upon users' perceived effectiveness and satisfaction. However, there was no
significant effect of efficiency showing that if we ignore the classification of users on
the basis of their flexible traveling behavior, the proposed and existing systems does
not influenced upon users' perceived efficiency.
6.3 Future Work
The computerized translation of customers' flexible and inflexible traveling behavior
into tailored and meaningful actionable results is currently not supported by any of the
reservation systems in place. Moreover, a lot of the process involved in preparing
tailor-made arrangements to cater for flexible and inflexible behavior of travelers can
be automated- or at the very least significantly be optimized through technology. The
current limitation ofthe research work is, a very basic Graphical User Interface design
for the proposed framework has been developed and implemented. However, for
giving flexible travel options early in the Online Airline Reservation Systems through
formal opaque quoting mechanism, a very fluid interface based on arbitrator system
between human input and database populated emails and documents is a potential
research area and it can further be standardised in a way which would work for lots of
traveling agencies and companies operating with different kinds of reservation
systems.
To proceed further, one can create a layer on top of the standard system -
everybody is booked onto a specific flight, but some days before the departure date
the additional layer looks for better options and reshuffles the passengers in the legacy
system below.
178
REFERENCES
[I] A. Schulz, "The Role of Global Computer Reservation Systems in the Travel
Industry Today and in the Future." Tourism, vol. 6. no. 2, pp. 17-20. 1996.
[2] D. Quinby, "Travel Agency Distribution Landscape: Key Terms and
Definitions," PhoCusWright Inc .. Sherman, CT 06784-1430 USA. Apr. 2008.
[3] B. Braham, "Computer Systems in the Hotel and Catering Industry," London:
Cassell, 1988.
[4] P. O'Connor, 'Tourism and Hospitality Electronic Distribution and
Information Technology," Oxford: CAB International, 1999.
[5] P. Sheldon, "Tourism Information Technology," Oxford: CAB International,
1977.
[6] The Transnational. (2008). Lufthansa, Swiss Add GDS Fees [Online].
Available: http:/ /WI'-w. thetransnational. tra ve I/ news. php? c i d= Lufthansa-
Swiss.Jan-08.24
[7] B. C. Smith, R. Darrow, J. Elieson, D. Gunther, B. V. Rao and F. Zouaoui,
"Travelocity Becomes a Travel Retailer," Interfaces, vol. 3 7, no. 1, pp. 68-81,
2007.
[8] Retail m Asia. (2010). Why Malaysia Tourism is Thriving [Online].
Available: http://www.retailinasia.com/article/markets/malaysia/20 I 0/1 0/why
malavs i a-tourism-thriving
[9] D. Quinby and R. Merten, 'The Role and Value of the Global Distribution
Systems in Travel Distribution," PhoCusWright Inc., Sherman, CT 06784-
1430 USA. Nov. 2009.
[10] K. Jerath, S. Netessine and S. K. Veeraraghavan, "Revenue Management with
Strategic Customers: Last-Minute Selling and Opaque Selling," Management
Science, vol. 56, no. 3, pp. 430--448, Mar. 2010.
[II] J. Sviokla, "Value Poaching," Across the Board (March/April 2003): 11-12.
179
[12] A. M. Ross, D. H. Rhodes, and D. E. Hastings, ''Defining System
Changeability: Reconciling Flexibility, Adaptability, Scalability, and
Robustness for Maintaining System Lifecycle Value,'' 17th
INCOSEinternational Symposium, San Diego CA, Jun. 2007.
[13] K. Jerath, S. Netessine and S. Veeraraghavan, "Selling to Strategic Customers:
Opaque Selling Strategies", Tepper School of Business, Paper 215, 2009.
http:/ /repository.cmu.edu/tepper/215
[14] K. Talluri and G. J. V. Ryzin, The Theory and Practice of Revenue
Management, Kluwer Academic Publishers, NY, 2004.
[15] M. Lambert. (2006). Last-Minute Gamble: Take a Chance, But Be Ready for
Anything [Online]. Available: http://www.accessmylibrary.com/article-1 G 1-
154985 891 /last-minutegam ble-take.html
[16] J. Harrison, Traveling Blind, Ottawa Citizen: L2, 2006.
[ 17] N. Granados, K. Han and D. Zhang, "Demand and Revenue Impacts of the
Opaque Channel: Empirical Evidence from the Airline Industry," Practice
Article, Dec. 3, 2010.
[18] S. Fay and J. Xie, "Probabilistic Goods: A Creative Way of Selling Products
and Services," Marketing Science, vol. 27, no. 4, pp. 674-690, 2008.
[19] D. Post, "Variable Opaque Products in the Airline Industry: A Tool to Fill the
Gaps and Increase Revenues," Journal of Revenue and Pricing Management.
Practice Article, 2010. www.palgrave-joumals.com/rpm/
[20] K. A. Olsen and A. Malizia, "Flexible User Interfaces for B2C Systems, Using
Booking as a Case Study," in Proceedings of the HCI 2005, Las Vegas, USA,
2007.
[21] A. Malizia, "Adding Flexibility to B2C Booking Systems Using a Virtual
Intermediate Travel Agent," in Proceedings of the IEEE Symposium on Visual
Languages and Human-Centric Computing, pp. 337-338, 2005.
[22] K. Charmaz, "Grounded Theory in the 21st Century," In Holstein & Gubrium.
Social Action Thousand Oaks, Cal: Sage, 2005.
[23] R. Miller, B. Johnson and W. W. Woolfolk, "Designing Flexible Information
Systems Saves Time and Money, Now and in the Future," Educause
Quarterly, no. 3, pp. 44-51,2002.
180
[24] N. C. Goodwin, "Functionality and Usability," Communications of the AClvf.
vol. 30,no.3,229-233, 1987.
[25] D. A. Norman, The Invisible Computer, Cambridge MA: MIT Press, 1998.
[26] P. Kalczynski. S. Senecal and J. Nantel. "Predicting Online Task Completion
with Clickstream Complexity Measures: A Graph-Based Approach."
International Journal of Electronic Commerce, vol. I 0, no. 3, pp. 123-143,
2006.
[27] E. Boling, "Usability Testing for Web Sites Learning for the Global
Community," in Seventh Annual Hypermedia Conference, Bloomington, 1995.
[28] M. Levi and F. Conrad, "A Heuristic Evaluation of a World Wide Web
Prototype," Interactions, vol. 3, no. 4, pp. 1-12, 1996.
(29] J. Nantel and S. Senecal, "The Effect of Counterproductive Time on Online
Task Completion," in M. Craig-Lees, G. Gregory and T. Davis (eds.).
Advances in Consumer Research Asia Pacific, 2007.
[30] J. W. Palmer, "Web Site Usability, Design, and Performance Metrics,"
Information Systems Research, vol. 13, no. 2, pp. 151-167,2002.
[31] J. Pitkow and C. Kehoe, "Emerging Trends in the WWW User Population,"
Communications ACM, vol. 39, no. 6, pp. I 06-108. 1996.
[32] J. Nielsen, Designing Web Usability, Indianapolis: New Riders Publishing,
2000.
[33] G. Saks. (2006). Travel: The Emergence of Meta Search [Online]. Available:
http://blog.compete.com/200611 1/14/meta-search-kayak -sidestep-farechase
mo bissim o-pinpoint-tra ve II
[34] S. Devaraj, M. Fan and R. Kohli, "Antecedents of B2C Channel Satisfaction
and Preference: Validating e-Commerce Metrics," Information Systems
Research, vol. 13, no. 3, pp. 316-331, 2003.
[35] D. McCloskey, "Evaluating Electronic Commerce Acceptance with the
Technology Acceptance Model," Journal of Computer Information Systems,
vol.44,no.2,pp.49-57,2004.
[36] T. P. Monsuwe, B. G. C. Dellaert and K. Ruyter, "What Drives Consumers to
Shop Online? A Literature Review," International Journal of Service Industry
Management, vol. 15, no. I, pp. I 02-121, 2004.
181
[37] M. Limayem, M. Khalifa and A. Frini, "What Makes Consumers Buy from
Internee A Longitudinal Study of Online Shopping," IEEE, vol. 30. no. 4,
2000.
[38] N. M. Shariff and M. Z. Shah, "Factors Influencing Travel Behavior and their
Potential Solution: A Review of Current Literatures", Jurnal A lam Bina, vol.
II, no.2, pp. 19-28, 2008.
[39] M. Manzari, "Intention to Use Internet Reservation Systems by Iranian Airline
Passengers," Master's Thesis, Lulea University of Technology, 2008.
[40] K. Jerath, S. Netessine and S. Veeraraghavan, "Selling to Strategic Customers:
Opaque Selling Strategies," in Consumer-Driven Demand and Operations
Management Models, Springer, pp. 253-300,2009.
[ 41] K. Jerath, S. Netessine and S. Veeraraghavan, "Selling to Strategic Customers:
Opaque Selling Strategies," in Consumer-Driven Demand and Operations
Management Models (eds. S. Netessine and C.S. Tang), pp. 253-300, 2009.
[42] J. Law, A Dictionary of Business and Management, Oxford University Press,
2009. http://www.oup.com/uk/isbn/9780199234899/.
[43] J. Browne. (1984). Classification of Flexible Manufacturing Systems. The
FMS Magazine. vol. 2, no. 2, pp. 114-117.
[44] A. K. Sethi and S. P. Sethi, "Flexibility in Manufacturing: A Survey," The
International Journal of Flexible Manufacturing Systems, vol. 2, pp. 289-328,
1990.
[45] R. E. Giachetti, L. D. Martinez, 0. A. Saenz and C. S. Chen, "Analysis of the
Structural Measures of Flexibility using a Measurement Theoretical
Framework," International Journal of Production Economics, vol. 86, pp. 4 7-
62,2003.
[46] J. H. Saleh, D. E. Hastings and D. J. Newman, "Flexibility in System Design
and Implications for Aerospace Systems," Acta Astronautica, vol. 53, no. 12,
pp.927-944,2003.
[47] A. M. King and S. Sivaloganathan, "Flexible Design: A Strategy for Design
Reuse," in Proceedings of Engineering Design Conference '98 on Design
Reuse, (Eds S. Sivaloganathan and T.M.M. Shahin), Professional Engineering
Publishing Limited, London.
182
(48] P. K. P. Rajan, M. VanWie. M. l. Campbell, K. L. Wood and K. N. Otto, "An
Empirical Foundation for Product Flexibility." /JcsiJin Studies. vol. 26, no. 4.
pp. 405-438. 2005.
149] J. Preece, Y. Rogers, H. Sharp. D. Benyon. S. Holland and T. Carey. Human
computer Interaction, Workingham, England: Addison- Wesley, 1994.
(50] R. Nilchiani, "Measuring Space Systems Flexibility: A Comprehensive Six
element Framework," PhD. Dissertation, Massachusetts Institute of
Technology: Department of Aeronautics and Astronautics, Cambridge, MA,
2005.
(51] A.M. Ross, "Managing Unarticulated Value: Changeability in Multi-Attribute
Tradespace Exploration," PhD. Dissertation, Massachusetts Institute of
Technology: Engineering Systems Division, Cambridge, MA, 2006.
[52] J. H. Saleh, "Weaving Time into System Architecture: New Perspectives on
Flexibility, Spacecraft Design Lifetime, and On-orbit Servicing," PhD.
Dissertation, Massachusetts Institute of Technology: Department of
Aeronautics and Astronautics, Cambridge, MA, 2002.
(53] R. L. Bryck, "Flexible Behavior Under Control 7 Neural and Behavioral
Evidence in Favor of a Two-component Model of Task-switching," PhD.
dissertation, Department of Psychology and the Graduate School, University
of Oregon, Eugene. 2008.
(54] P. Barnard, "The Contributions of Applied Cognitive Psychology Toward the
Study of Human-computer Interaction," 1995. In R. M. Baecker, J. Grudin,
W.A.S. Buxton and S. Greenberg (eds.), Readings in Human-Computer
Interaction: Toward the Year 2000. Second Edition (pp. 640-658). San
Francisco, CA:Morgan Kaufman Publishers.
(55] S. K. Card, T. P. Moran and A. Newell, The Psychology of Human-Computer
Interaction, Hillsday, J: Lawrence Erlbaum, 1983.
(56] T. R. G. Green, S. P. Davies and D. J. Gilmore, "Delivering Cognitive
Psychology to HCI: The Problems of Common Language and of Knowledge
Transfer," Interactin}i with Computers, vol. 8, pp. 89-111, 1996.
[57] T. K. Landauer, "Let's Get Real: A Position Paper on the Role of Cognitive
Psychology in the Design of Humanly Usefid and Usable Systems," 1995. In
R.M. Baecker, J. Grudin, W.A.S. Buxton and S. Greenberg (eds.), Readings in
183
Human-Computer Interaction: Toward the Year 2000. Second Edition (pp.
659--665). San Francisco, CA: Morgan Kaufman Publishers.
[58] E. Pollitzer and E. Edmonds, "Editorial: The Evolving Partnership between
Cognitive Science and HCI," International Journal of Human-Computer
Studies, vol. 44, pp.731-741, 1996.
[59] R. Guindon, Cognitive Science and its Applications for Human-Computer
Interaction, Hillsdale, NJ: Lawrence Erlbaum, 1988.
[60] M. M. Gardiner and B. Christie, Applying Cogntiive Poychologv to User
Interjiace Design, Chichester, UK: John Wiley & Sons, 1997.
[61] K. E. Flaherty and J. C. Mowen. (2010). The Role of Schemas in Consumer
Behavior Research [Online]. Available:
http:/ /onlinelibrary. wiley.com/doi/1 0.1002/9781444 316568.wiem03055/abstra
ct
[62] A. D. Palma and D. Rochat, "Understanding Individual Travel Decisions:
Results from a Commuters Survey in Geneva," Transportation, vol. 26, pp.
263-281, 1999.
[63] J. E. Mills and R. Law, "Handbook of Consumer Behavior, Tourism, and the
Internet," Journal of Travel and Tourism Marketing, vol. 17, no. 2/3, 2004.
[64] M. Sigala, A. Lockwood and P. Jones, "Strategic Implementation and IT:
Gaining Competitive Advantage from the Hotel Reservation Process,"
International Journal of Contemporary Hospitality Management, vol. 17, no.
3,pp.364-371,2001.
[65] W. D. Walls, "Competition in the Market for Travel Information Services,"
Proceedings of the Eastern Asia Society for Transportation Studies, val. 5, pp.
533-544, 2005.
[ 66] Orbitz. (20 1 0). Flights [Online]. Available:
http:/ /www.orbitz.com/ App/PrepareFlightsTab
[67] I. H. Hann and C. Terwiesch, "Measuring the Frictional Costs of Online
Transactions: The Case of a Name-Your-Own-Price Channel," Management
Science, vol. 49, no. 11, pp. 1563-1579,2003.
[68] C. Terwiesch, S. Savin and I. H. Hann, "Online Haggling at a Name-Your
Own-Price Retailer: Theory and Application," Management Science, vol. 51,
no.3,pp.339-351,2005.
184
[69 J Priceline. (201 0). Name Your Own Price [Online]. Available:
http:! /www. price! ine.com/
[70] Hotwire. (2010). Welcome to Flexible Date Search [Online]. Available:
http://www.hotwire.com/uslindex.jsp
[71] A. Hill, 'Top 5 Reasons Your Customers Abandon Their Shopping Carts," Zijf
Davis Smart Business for the New Economy, vol. 14, no. 3, pp. 80-85, 2001.
[72] L. R. Klein, "Evaluating the Potential of Interactive Media through a New
Lens: Search vs. Experience Goods," Journal of Business Research, vol. 41,
pp. 195-203, 1998.
[73] W. 1. Doll, A. Hendrickson and X. Deng, "Using Davis's Perceived
Usefulness and Easeof-use Instruments for Decision Making: A Confirmatory
and Multigroup Invariance Analysis," Decision Sciences, vol. 29, no. 4, pp.
839-869, 1998.
[74] J. Nielsen, Designing Web Usability: The Practice a/Simplicity, Indianapolis:
New Riders Publishing, 2000.
[75] J. Nielsen, Homepage Usability: 50 Websites Deconstructed, Indianapolis:
New Riders, 2002.
[76] A. 1. Head, Design Wise: A Guide for Evaluating the Interface Design of
Information Resources, New Jersey: CyberAge books, 2000.
[77] S. Krug, Don't Make me Think!, Indianapolis: New Riders Publishing, 2000.
[78] R. Williams and J. Tollett, The Non-designers Web Book. Berkeley, CA:
Peachpit Press, 2000.
[79] F. D. Davis, "A Technology Acceptance Model for Empirically Testing New
End-User Information Systems: Theory and Results," PhD. dissertation, MIT
Sloan School of Management, Cambridge, MA, 1986.
[80] N. E. Miller, and J. Dollard, Social Learning and Imitation, New Haven: Yale
University Press, 1941.
[81] E. M. Rogers, Diffusion of Innovations, 5th ed., New York, NY: Free Press,
2003.
[82] M. Fishbein and I. Ajzen, Belief,' Attitude, Intention and Behavior: An
Introduction to Theory and Research, Reading, MA: Addison-Wesley, 1975.
185
[83] I. Ajzen and M. Fishbein, "Attitude-Behavior Relations: A Theoretical
Analysis and Review of Empirical Research," P.1ychological Bulletin, vol. 84,
pp. 888-918,1977.
[84] I. Ajzen and M. Fishbein, Understanding Attitudes and Predicting Social
Behavior, Englewood Cliffs, NJ: Prentice Hall, 1980.
[85] H. C. Triandis, Interpersonal Behavior, Monterey, Calif.: Brooks/Cole Pub.
Co, 1977.
[86] P. Legris, J. Ingham and P. Collerette, "Why Do People Use Information
Technology? A Critical Review of the Technology Acceptance Model,"
Information & Management, vol. 40, pp.l91-204, 2003.
[87] R. 0. Briggs, M. Adkins, D. Mittleman, J. Kruse, S. Miller and J. F.
Nunamaker, "A Technology Transition Model Derived From Field
Investigation of GSS Use Abroad the U.S.S. CORONADO," Journal of
Management Information Systems, vol. 15, no. 3, pp. 151-195, 1999.
[88] D. S. Roberston, "Social Determinants of Information System Use," Journal
of Management Information Systems, vol. 5, no. 4, pp. 55-71, 1989.
[89] D. Graber, Processing the News: How People Tame the Information Tide, 2nd
ed., New York: Longman, pp. 119 & 133, 1988.
[90] Y. Lee, K. A. Kozar and K. R. T. Larsen, "The Technology Acceptance
Model: Past, Present, and Future," Communications of the Association for
Information Systems, vol. 12, no. 50, pp. 752-780,2003.
[91] H. C. Lucas and V. K. Spitler, "Technology Use and Performance: A Field
Study of Broker Workstations," Decision Sciences, vol. 30, no. 2, pp. 291-
311, 1999.
[92] V. Venkatesh and F. Davis, "A Theoretical Extension of the Technology
Acceptance Model: Four Longitudinal Field Studies," Management Science,
vol.46,no.2,pp. 186-204,2000.
[93] D. Robey, "Research Commentary: Diversity in Information Systems
Research: Threat, Promise, and Responsibility," Information Systems
Research, vol. 7, pp. 400-408, 1996.
[94] Y. Malhotra and D. F. Galletta, "Extending the Technology Acceptance Model
to Account for Social Influence: Theoretical Bases and Empirical Validation,"
186
m Proceedings or the 32nd I /awaii !nternalional C 'onference on System
Sciences. Hawaii, 1999.
[95] M. Y. Chuttur, "Overview of the Technology Acceptance Model: Origins,
Developments and Future Directions." Indiana University, USA. Sprouts:
Working Papers on Information Systems. vol. 9. no. 37, 2009.
http://sprouts.aisnet.org/9-3 7
[96] V. Venkatesh, "Determinants of Perceived Ease of Use Integrating Control,
Intrinsic Motivation, and Emotion into the Technology Acceptance Model,"
Information Systems Research. vol. II, no. 4, pp.342~ 365, 2000.
[97] P. Benckendorff, "An Exploratory Analysis of Traveler Preferences for
Airline Website Content," Jnrormation Technology & Tourism, vol. 8, pp.
149~159, 2006.
[98] S. Balgolu and Y. A. Pekcan, "The Website Design and Internet Site
Marketing Practices of Upscale and Luxury Hotels m Turkey," Tourism
Management, vol. 27, no. I, pp. 171~176, 2006.
[99] P. 1. Benckendorff and N. L. Black, "Destination Marketing on the Internet: A
Case Study of Australian Regional Tourism Authorities," Journal of Tourism
Studies, vol. I], no. I, pp. II ~21, 2000.
[100] L. Cai, J. A. Card and S. T. Cole, "Content Delivery Performance of World
Wide Web Site of US Tour Operators Focusing on Destination in China,"
Tourism Management, vol. 25, pp. 219~227. 2004.
[101] R. Law and R. Leung, "A Study of Airlines' Online Reservation Services on
the Internet," Journal of Travel Research, vol. 39, pp. 202~211, 2000.
[I 02] 1. Murphy, E. J. Forrest, C. E. Wotring and R. A. Brymer, "Hotel Management
and Marketing on the Internet," Cornell Hotel and Restaurant Administration
Quarterly, vol. 37, no. I, pp. 61~71, 1996.
[103] C. Wan, "The Web Sites oflnternational Tourist Hotels and Tour Wholesalers
in Taiwan," Tourism Management, vol. 23, pp. !55~ 160, 2002.
[104] A.M. Morrison, J. S. Taylor and A. Douglas, "Website Evaluation in Tourism
and Hospitality: The Art is Not Yet Stated," Journal of Travel and Tourism
Marketing, vol. 17, no. 2/3, pp. 233~251, 2004.
187
[105] A. Schar!, K. W. Wober and C. Bauer, "An Integrated Approach to Measure
Web Site Effectiveness in the European Hotel Industry," Information
Technology & Tourism, vol. 6, pp. 257-271,2004.
[106] J. A. Siguaw, C. A. Enz and K. Namasivayam, "Adoption of Information
Technology in US Hotels: Strategically Driven Objectives," Journal of Travel
Research, vol. 39, pp. 192-201,2000.
[107] Y. Yuan, U. Gretzel and D. R. Fesenmaier, "Internet Technology Use by
American Convention and Visitors Bureaus," Journal of Travel Research, vol.
4l,pp.240-255,2003.
[108] H. Nysveen, L. B. Methlie and P. E. Pedersen, "Tourism Websites and Value
Added Services: The Gap between Customer Preferences and Web Sites'
Offerings," Information Technology & Tourism, vol. 5, pp. 165-174, 2003.
[109] J. A. Card, C. Chen and S. T. Cole,. "Online Travel Products Shopping:
Differences between Shoppers and Non-shoppers," Journal of Travel
Research, vol. 42, pp. 133-139. 2003.
[110] A. M. Susskind, M. A. Bonn and C. S. Dev, "To Look or Book: An
Examination of Consumers' Apprehensiveness toward Internet Use," Journal
of Travel Research, vol. 41, pp. 256-264, 2003.
[Ill] S. Jang, "The Past, Present and Future Research of Online Information
Search," Journal of Travel and Tourism Marketing, vol. 17, no. 2/3, pp. 41-
47,2004.
[112] M. Luo, R. Feng and C. Liping, "Information Search Behavior and Tourist
Characteristics: The Internet vis-a-vis other Information Sources," Journal of
Travel and Tourism Marketing, vol. 17, no. 2/3, pp. 15-25,2004.
[113] B. Bai, C. Hu, J. Elsworth and C. Countryman, "Online Travel Planning and
College Students: The Spring Break Experience," Journal of Travel and
Tourism Marketing, vol. 17, no. 2/3, pp. 79-91,2004.
[114] A. Rudstrom and P. Fagerberg, "Socially Enhanced Travel Booking: A Case
Study," Information Technology & Tourism, vol. 6, pp. 211-221,2004.
[115] S. Klein, F. Kohne and A. Oorni, "Barriers to Online Booking of Schedule
Airline Tickets," Journal of Travel and Tourism Marketing, vol. 17, no. 2/3,
pp. 27-39,2004.
188
[ 116] C. H. Marcussen. "The Effects of Internet Distribution of Travel and Tourism
Services on the Marketing Mix: No-frills. Fair Fares, and Fare Wars in the
Air," Information Technolor;y & Tourism. vol. 2. pp. 197-212. 1999.
[117] G. Gallego and R. Phillips. "Revenue Management of Flexible Products,··
Manufac/llring and Service Operations Management, vol. 6, no. 4, pp. 321-
337. 2004.
[118] Y. Jiang, "Price Discrimination with Opaque Products," Journal of Revenue
and Pricing Management, vol. 6, no. 2, pp. 118-134,2007.
[119] R. J. Deneckere and R. P. McAfee, "Damaged Goods," Journal of Economics
and Management Strategy, vol. 5, pp. 149-174, 1996.
[120] X. Huang and G. Sosic, "Name-Your-Own-Price as a Competitive Secondary
Channel in the Presence of Posted Prices," Working Paper. Concordia
University, Montreal, Canada, 2010.
[121] S. Fay, "Selling an Opaque Product through an Intermediary: The Case of
Disguising One's Product," Journal of Retailing, vol. 84. no. I, pp. 59-75,
2008.
[122] N. Granados, A. Gupta and R. J. Kauffman, "Designing Online Selling
Mechanisms: Transparency Levels and Prices," Decision Support Systems,
vol. 45, no. 4, pp. 729-745, 2008.
[123] S. L. Puller, A. Sengupta and S. N. Wiggins, "Testing Theories of Scarcity
Pricing and Price Dispersion in the Airline Industry." Working Paper, Texas
A&M University, 2008.
[124] N. Granados, A. Gupta and R. J. Kauffman, "Information Transparency in
Business-to-Consumer Markets: Concepts, Framework, and Research
Agenda." Information Systems Research, vol. 21, no. 2, pp. 207-226.2010.
[125] D. Shapiro and X. Shi, "Market Segmentation: The Role of Opaque Travel
Agencies," Journal of' Economics & Management Strategy, vol. 17, no. 4, pp.
803-837. 2008.
[126] G. Ng-Kruelle, P. Swatman and 0. Kruelle, "e-Ticketing Strategy and
Implementation in an Open Access System: The case of Deutsche Bahn,"
2006. http:/ 1\V\\<w.insv I. unisa.edu.au/publications/work ing -papers/2006-08. pdf
[127] J.P. Schwieterman, "Fare is Fair in Airline Deregulation: The Decline of Price
Discrimination," Regulation, vol. 9, no. 4. pp. 32-38. 1985.
189
[128] L. /\. Garrow, ''Online Travel Data: A Goldmine of New Opportunities."
.Journal of Revenue and Pricing Management, vol. 8, no. 2, pp. 24 7-254.
2009.
[129] W. J. Carroll. R. J. Kwortnik and N. L. Rose, "Travel packaging: An Internet
Frontier," Cornell Ho,pitality Report, vol. 7, no. 17. pp. 4-16.2007.
[130] J. Doerr, D. Kerkow, T. Koenig, T. Olsson and T. Suzuki, "Non-functional
Requirements in Industry - Three Case Studies Adopting an Experience-based
NFR Method," in Proceedings of the 2005 13th IEEE International
Conference on Requirements Engineering (RE'05), Washington, DC, USA,
pp. 373-384.
[131] Y. Y. Thaw, A. K. Mahmood and P. D. D. Dominic, "A Study on the Factors
that Influence the Consumers' Trust onE-Commerce Adoption," International
Journal of Computer Science and Information Security (IJCSIS), vol. 4, no. 1
and 2, pp. 153-159, 2009.
[132) J. P. Chin, V. A. Diehl and K. L. Norman, "Development of an Instrument
Measuring User Satisfaction of the Human-computer Interface," in
Proceedings of the SIGCHI Conference on Human Factors in Computing
Systems CHI '88, ACM, New York, NY, USA, 1988, pp. 213-218.
[133] H. X. Lin, Y. Y. Choong and G. Salvendy, "A Proposed Index of Usability: A
Method for Comparing the Relative Usability of Different Software Systems,"
Journal of Behaviour and Information Technology, vol. 16, no. 4/5, pp. 267-
278, 1997.
[134] J. R. Lewis, "IBM Computer Usability Satisfaction Questionnaires:
Psychometric Evaluation and Instructions for Use," International Journal of
Human-Computer Interaction, vol. 7, no. 1, pp. 57-78, 1995.
[135] N. Burns, S. Grove, "Introduction to Quantitative Research," in The Practice
of Nursing Research: Conduct, Critique, and Utilization, 2nd ed. Philadelphia:
WB Saunders Company, 1993.
[136) C. F. Waltz, R. B. Bausell, "Quantitative Research," in Nursing Research:
Design, Statistics and Computer Analysis, 2nd ed. Philadelphia: FA Davis
Company; 1983.
[137) J. Aldrich, "Correlations Genuine and Spurious in Pearson and Yule,"
Statistical Science, vol. 10, no. 4, pp. 364-376, 1995.
190
[138] J. P. Chin, V. A. Diehl and K. L Norman, ''Development of an Instrument
Measuring User Satisfaction of the Human-Computer Interface," in CHI '88
Proceedings of the SIGCHI conference on Human factors in computing
systems, pp. 213-218.
[139] H. X. Lin, Y. Y. Choong and G. Salvendy, ''A Proposed Index of Usability: A
Method for Comparing the Relative Usability of Different Software Systems,''
Behavior and Information Technology, vol. 16, no. 4/5, pp. 267-278, 1997.
[140] P. Donald, "Behavioral Criterion," in Essential Personality. Hodder Arnold,
2003.
[141] S. A. Mcintire and L. A. Miller. Foundations of Psychological Testing, 2nd
ed. Thousand Oaks, CA: Sage. 2005.
[142] 0. Johnson, "About the Information Theory and the Central Limit Theorem,"
in Information Theory and the Central Limit Theorem. Imperial College Press,
2004, p. 88. ISBN 1860944736.
[143] S. Wright, "Effective Size of Population," in Evolution and the Genetics of
Populations, 2nd ed. Chicago: University of Chicago Press, 1984. ISBN 0-
226-91039-3.
[144] M. S. Rahmat. (2000). Grounded Theory Methodology as the Research
Strategy for a Developing Country [Online]. Available: http://nbn
resolving.de/urn:nbn:de:O 114-fgsOOO 1198
[145] A. Strauss and J. Corbin. Basics of Qualitative Research: Grounded Theory
Procedures and Techniques. Newbury Park, CA: Sage, 1990.
[146] B. G. Glaser and A. L. Strauss, "Grounded Theory," in The Discovery of
Grounded Theory. Strategies for Qualitative Research. New York: Aldine
Pub. Co, !967, p.27l.
[147] R. W. Riley and L. L. Love, "The State of Qualitative Tourism Research,"
Annals of Tourism Research, vol. 27, no. 1, pp. 164-187, 2000.
[148] K. Charmaz, "Grounded Theory in the 21st Century." In Holstein & Guhrium.
Social Action Thousand Oaks, Cal: Sage, 2005.
[149] Marshall, Catherine and G. B. Rossman, Designing Qualitative Research, 3rd
Ed. London: Sage Publications. pp. 115-1999.
191
[ 150] A. M. Schon1and and P. W. Williams, "Using the Internet for Travel and
Tourism Survey Research: Experiences from the Net Traveller Survey,"
Journal of Travel Research, vol. 35, no. 2, 1996.
[151] A. Decrop, "Triangulation in Qualitative Tourism Research," Tourism
Management, vol. 20, no. 1, pp. 157-161, I999.
[152] D. K. Padgett, Qualitative Methods in Social Work Research: Challenges and
Rewards. Thousand Oaks, CA: Sage, I 998.
[153] M. Q. Patton, Qualitative Evaluation Methods. Beverly Hills, CA: Sage, 1980.
[154] N. Kowtha and Choon, "Determinants of Website Development: A Study of
Electronic Commerce in Singapore," Information & Management. vol. 39, no.
3,pp.227-242,2000.
[155] B. D. Wright, "IRT m the 1990s: Which Models Work Best?" Rasch
Measurement Transactions, vol. 6, no. 1, pp. 196-200, 1992.
[156] A. Field. (2005). Project 3: Design a Questionnaire [Online]. Available:
http://coe.georgiasouthern.edu/foundations/bwgriffin/edur9131/ Andy Field re
search project.pdf
[157] J. P. Ferrando, "Difficulty, Discrimination and Information Indices in the
Linear Factor-Analytic Model for Continuous Responses, Applied
Psychological Measurement, vol. 33, pp. 9-24, 2009.
[158] Kelley
[159] HHS, & GSA. (2006). Research-based Web design and usability guidelines
[Online]. Available: http://usability.gov/pdfs/guidelines.html
[160] J. Nielsen and R. L. Mack, Usability lmpection Methods (Eds.). New York,
NY: John Wiley & Sons, 1994.
[161] N. Brace, R. Kemp and R. Snelgar, "Multiple Regression," m SPSS for
Psychologists, 3rd ed. Mahwah, NJ: Lawrence Erblaum, 2006.
[162] Laerd Statistics. (2011). Statistics [Online]. Available:
http:// sta tis ti cs.laerd. corn/ statistical-guides/ descriptive-inferential-
statistics. php
[163] J. L. Rodgers and W. A. Nicewander, "Thirteen Ways to Look at the
Correlation Coefficient," The American Statistician, vol. 42, no. 1, pp. 59-66,
Feb. 1988.
192
[164 J M. S. Stigler, ''Francis Galton's Account of the Invention of Correlation,"
Statisrical Science, vol. 4. no. 2. pp. 73-79. 1989.
[165] W. Q. Meeker and L. A. Escobar. "Reliability Concepts and Reliability Data,"
in Sratisrical Methods for Reliahi/iry /)ala. Hoboken. New Jersey: Wiley,
1998.
[166] J. M. Cortina, "What Is Coefficient Alpha'> An Examination of Theory and
Applications," Journal of Applied P!>ychology, vol. 78, no. 1, pp. 98-104,
1993.
[167] StatSoft (2011). Reliability and Item Analysis [Online]. Available:
http :1 /v.ww. statso ft. com/textboo kl re l iabi I i tv -and- i tern -analysis/
[168] StatSoft. (2011). ANOV A/MANOV A [Online]. Available:
http://v.ww.statsoft.com/textbooklanova-manova/
[169] A. Agresti. (2007). Chapter 12 Comparing Groups: Analysis of Variance
(ANOVA) Methods [Online]. Available:
http:/ /www.stat.ufl.edu/-aa/sta6 I 27 /ch 12.p<jf
[170] MEDCALC. (2011). Analysis of Covariance [Online]. Available:
http: I lwww .m edcalc. org/man ual/ anal vsis-o f-co variance. ph p
[171] R. Lowry. (2011). Chapter 17: One-Way Analysis of Covariance for
Independent Samples Part 2 [Online]. Available:
http:/ /faculty.vassar.edu!lowry/ch 17pP .html
[172] Statistically Significant (2011 ). MANOVA [Online]. Available:
http://www. statisticall ysi gnifi cantcons ul ting. com/Manova-Statistic-S P S S .htm
[173] G. Carey. (1998). Multivariate Analysis of Variance (MANOVA): I. Theory
[Online]. Available:
http://ibgv,ww.colorado.edu/-carey/p 7291 dir/handouts/manova !.pdf
[174] G. G. David. (2011). Multivariate GLM, MANOVA, and MANCOVA
[Online]. Available: http://facultv.chass.ncsu.edu/garson/P A 765/manova.htm
[175] UCLA: Academic Technology Services. (2011). Stata Annotated Output -
MANOVA. Available:
http://www.ats.ucla.edu/stat/stata/output/Stata MANOVA.htm
[176] J. P. Stevens, "Multivariate Analysis of Variance," in Applied Multivariate
Statistics for the Social Sciences, 41h ed. Mahwah. NJ: Lawrence Erblaum,
2002.
193
[177] F. Galton, "Kinship and Correlation," in Statistical Science. vol. 4, no. 2, pp.
80-86, 1989. http://www.jstor.org/stable/2245330
[178) G. U. Yule, "On the Theory of Correlation," Journal of the Royal Statistical
Society. vol. 60, no. 4, pp. 812-54, 1987.
[179) K. Pearson, G. U. Yule, N. Blanchard and A. Lee, "The Law of Ancestral
Heredity," Biometrika, vol. 2, no. 2, pp. 211-236, 1903.
[180) MEDCALC. (2011). Multiple Regression [Online). Available:
http://www.medcalc.org/manual/multiple regression. php
[181) Stevens. (1999). Post Hoc Tests in ANOVA [Online]. Available:
http://pages.uoregon.edu/stevensj/posthoc.pdf
[182) N. Crichton, "Methodological Issues in Clinical Research," Journal of Clinical
Nursing, vol. 10, pp. 707-715, 2001.
[183) UNESCO. (2011). Non-parametric Measures of Bivariate Relationships:
Kendall's Tau [Online]. Available:
http://www. unesco.org/we bworld/idams/advguide/Chapt4 2.htm
[185) Skeeterssays. (2008). On the Use of Error Bars [Online). Available:
http:// skeetersays. b I o gspot.com/2008/0 7/ on-use-of-error -bars. htm I
[186) JMP. (2008). What Good are Error Bars? [Online). Available:
http://blogs.sas.com/j mp/index.php? /arc:hives/127-What -Good-Are-Error
Bars.html
[187) J. Pitocchelli. (2001). Multiple Comparisons Tests [Online]. Available:
http://www.anselm.edu/homepage/jpitocch/anovalmultcomp.html
[188) Airport International. (2011). How Important is Ticket Flexibility Really
Necessary for Business Travellers? [Online). Available: http://www.airport
int.com/ arti c I e/how-important-is-ticket-flexi bi I i ty-real! y -necessary-for
business-travellers.html
[189) G. Pel!. (2010). Top 5 Reasons Flights Get Canceled [Online). Available:
http://www.splendora.com/tive things on my radar by splendora founder
gina pell/5 top 5 reasons flights gel canceled
194
LIST OF PUBLICATIONS
Journal Publications
1. A. Mushtaq, S. Sulaiman, P. D. D. Dominic, S. Riaz and M.A. T. Khalil, "A
Study to Examine If Integration of OTAs Features can make SBTs more
Flexible Online Airline Reservation Systems." Information Technology
Journal, vol. JO,no. 7,pp.I315-1324, 2011.
2. A. Mushtaq, S. Sulaiman and S. Riaz. "Flexibility Concept: Facilitating Post
Sale Changes in Online Reservation Systems," International Journal on
C 'omputer Science and Engineering (IJCSE}, vol. 1. no. 3, pp. 156-162. 2009.
Conference Punlications
1. A. Mushtaq, S. Sulaiman, P. D. D. Dominic and S. Riaz, "A Qualitative
Enquiry to Explore the Concept of Users' Perception on Factors Influencing
Flexible Traveling Behavior and Flexible Online Airline Reservation
Systems," in Proceedings of the Digital Enterprise and Information Systems,
E. Ariwa and E. El-Qawasmeh (Eds.): DEIS 201 I. CCIS 194, pp. 739-753,
Springer-Verlag Berlin Heidelberg 2011.
2. A. Mushtaq, S. Sulaiman and J. Loviscach, "Issues with Flexibility: A Reason
for not Using Online Reservation Systems?" in Proceedings of the IEEE
International Symposium on Information Technolozy, vol. 3, pp. 1076-1081,
2010.
3. A. Mushtaq, S. Sulaiman, P. D. D. Dominic and S. Riaz, "Green Environment
with Flexible Airline Reservation Systems." in Proceedings of the IEEE
International Conference on Intelligence and Information Technology,
Lahore, vol. 1, pp. 488-492, 2010.
195
APPENDIX A
SURVEY ON EVALUATION OF ONLINE AIRLINE RESERVATION SYSTEMS
SL'RVEY 0:\" EYALL'ATIO:\" OF 0:\"LDIE AIRLDIE RESERVATIOX SYSTE:\IS
Objecthe- The objective of this- s.urYey is to find out physiological, cognitive sccial behavior and the experience of the selected group of people -;vho used airline online res.en:ation systems.
The feedback ofpeople may help to propose a better framework \'Vhich could be used to design more flexible online reser.,:ation systems.
Last date of surYe)· fonn submission- 15 October 2009
::"ote- Please return. the fonn to the person·place from >.\here it was. collected before or by the deadline In cas.e of queries, please do not hesitate to contact the tmdenigned either by call or email .
. .\.rif~Iusl1taq PhD Scholar Department ofComputer and Information Sciences. Universiti Teknologi PETRONAS Bandar Seri Iskandar,31 750 Tronoh Perak, :\lalays.la -.-{10195969598 coolafeellvahoo.com
SECTION A- YOCR DETAILS (Pleas~ place a ri.ck (i) at the appropriate box)
1. Age
Gender
3. ~ationality
flll6-c5 QJ 26-35
f3l 36-45
8J 46-55
m Fernale
[3] Palci3.tani
[5l 56-65
(I] Abov<!6:5
[Jl Malavsian
Q EU~ational [TI Others (specify-----------
j
6
Educational background
How frequently you shop online?
Ha:\'e you ever booked an online airline ticket?
Ill Post-graduate
OJ r..1aster
OJ ~fosth· OJ Occastonally
f3l Under-graduate
QJ Se-:ondary
[iJ "\~es r-if --~--~~"pl-ease at1:empt SECTI~~ B only)
ITJ ~o (tf ··_No-·pleHe anemptSECTIO~ Conly)
SECTIO:'ol B- L'SERS WITH OXL!XE EXPERIE:'<CES (Pluu pJa.ce a tick (I) at the appropriate box)
1. \l.lrich of the follo¥.1ng airline's web-site youhaveused to book ticket':'
TRAVELLED Name of 1h-t: Airline Lastslx Lastont'
months '\'ear 1. ~falaysia Airline 2. Emirates Air
196
More than oaev~.ars
6
Briti~h .-1.J.r.':a'\ · s Tiu1.-\irwa'.'S Lufl:hama .-illlme Pakntan Intemational.~line .-\it .-\~ia
S Other (please specif\ name)
Flexibilhy of Existing Systems
Ha\·e you eYertried to nuke changes in your traveling dates online"
3 Ha•,:e you e\'ertned to cancehour ticket onlme''
-1. Hanyouevertried to transferyomticket to someom else online~
5 Hanyoue,·ertried to conect error (such ai- typos)in you name orin yom address online'?
6 Ha·•eyoue,·ertried to res.ern a t1cket for fe,-..· dayn;,ith the intention to buv it later"
NeTI!r t:"aable, ro Option 11D1
Sue<eUful do .. naibb]e
Rank the follOt\ing feature.i from 1 to 5 (lmo~t important- 5 least important) you \>:ouldlike to me in online svs.tem
Rank Rank
Online changes in traveling schedule Online correction oft:vpo errors.
Online cancellationofticket Onhne reierYationofticket
Online transftr ofncket
Please indicate (tick) your len I of agreement to the staternent belo"
Strongly Disagree {SDA) Disagree {DA) Strongly Agree (SA)
Evaluation of Existing Onl.he Systems
S. Reading characters on the screen of online reservation ,..,-eb site was hard
9 Organization ofinformation was confusing
1 0_ Sequence of screens. was confusing_
11 Use of terms tluoughout system \\a:> inconsistent
12. Position of messages on screen ~vas inconsistent
13 Learrungto operate the system '1.\".:lS difficult
1-1.. P:ro,·ides effective linkage '-\ith othertravel-related panners (e g. link to other airline resen·ation s.y:.tem in case of connected flights)
15. The design for data entry \vas flexible
16. The-.•:eb site proYtdes t1exible user guidance.
197
SllA DA N A
' 4
4
4
4
4
4
4
4
""
' -
1- Completionofprocessing "''·as clea1ly indicated.
18 It requires the fe.·.~:est :;teps po:>s.tble. to accomplish ',':hat I ·.nnt to do •;1th it.
19. I can u:;e it without reading imtruct10m
20. Both occasional and regular users would bke it
21_ I canreco•·er from mistakes quickly and eailly
"'"' I \YOuld prefer to me online :;ystem for future bookmg.
23_ h:ouldreconunendit to a friend.
2-t. 0\·erall, I am sati.;fied '-"i.th the support information (online.hne help, messages, documentation) •• .,hen completing the tasks.
25. Overall. I am satisfied '-"i.th the ease of completing the tasks in this scenario.
26. O.·erall I am satisfied \;.ith the amormt oftime it took to complete the tasks in this scenario.
SECTIO::-< C- l'SERS WITHOl.'T 0:-<LI="E EXPERIE~CES Plus~ indicate (tick) your level of agr~ement to the statement below
;
Disagee (DA) Sttongly Agee (SA) 5
Reason for uot Using Oaline- Reun•tion System'S SDA DA s
1. I do not haYe access to internet
0 I do not kno\Y ho'"·to book an online ticket. 2 ;
3. I am not comfortable shopplngonline.
4. I do not have credit·debit card to make an online booking
5. I do not .,...-ant to trv because chances. of frauds are there.
6. Online resel\.:ation systems. are very complicated to use.
I do not use online :;ys.tems because if some mistake "'ill O(cur, how 'i\·ould I correct it':'
:
S. I prefer to go to tranl agent because they can find a better package forme.
9. I prefer to go to tra,·el agent because they are more reliable. ' 10. I prefer to go to travel agent becallie I can request them to make change:. in my : ;
flight schedule. 11. I prefer to go to travel agent because I canreser·:e tickets {for few days)
' \\i.thouteven pavlnga pennv.
A SA
" 5
" 5
4
" 4
4
" 4
4
" 5
4
12. Rank the 3 mo:>t important reasons (from the above statements 1 to 11) for not using online systems 1._ 2. 3
~Thank you \'ety much for completing this questiotu1aire :;;::
198
APPENDIX B
USERS' FLEXIBLE BEHAVIOR IN TERMS OF COMPROMISING ON SERVICE
QUALITY ATTRIBUTES
Csers' Flexible Behavior in term of Compromising on Sen·ice QualitY Attributes
Objecth-i!- The objecti\'e of this questionnaire is to fmd out the percentage of flexible usen that "'ill s:uppott the flexibility ofrese:rvation systems The feedback recelYed 1nay help in desigrung znore flexible reservation system;;
I. Age
Gender
General Informatiou fP(,.aH pla~,.,. rUk () ar th,. appropt'iaor" box I
rn Bela,_,,. 20 yean rn .::I to .30 years
~ 1 to ::'0 years - 51 to 60 years
[iJ l\1ale D Female
31 to 40 years
:\bm·e 60 years
E ducat.ional Backg:rm.md rn H1gh ~chool [}]Diploma ~Gr-aduate Post-graduate
-! Occupation rn Student Q Profes~onal [J] Self Employed
Renred
' Use of Online Tool [iJ SBT D OTA
Ple:ue indicate which of the following statements describes you in the but possib1t- way'!
Nilitiil
Identity
6. I am not a frequent flyer, I fly only once m a year or w
I normally fly to spend holiday~ ~vith fantily friends
S I 01mnot a status conscious person_ I can ily using any class
~f-efficaey
9 I an1. excellent in computer skill:;~ I can book my itin<:"rary in a-ver..· ilion time
10 I am a regulat· ne·ws reader: I a.h\·ays ke<"p myselfupdated on totmrm
11 I am e:xcellent in managing trips. I c.:~.nplanm:v trip in a very ilion tune notico;:
1: .. >\:irline ·s J"epute does not fascinate me a lot. it plays a tninOJ" role in m,reservations
13. I do not alwavs prefer mv own countn· au-hne while res<'!!''Ul.g my tickets
1-t I do not ahvay$ prefer to fl"\ dwlng occastoru
1 :5 I do not alway~ prefer to flv from nr; O',\ll to•.\n_ I c'l.n take 4 tbght from mv neighbor to~·.n
16 I do not ahva~, s prefer 4irlin<'!s with better leg space ar,..a and comfonable seats
1- I do not ahvays prefeJ" arrlines wluch pro\ tdf: Jet ways, so that I don I haYe to clitnb the gan_s;~·way and wait for bu:;se $
199
- - 0
- - 0
- - 0
+ ++
+ ++
+ ++
Engagement
1 S. The a \·ailability of rewards from customer loyalty programs motiYates.meto us.ethe same airline for my next journey.
19. Staff atlitude not ahva ~·smotintesme to me the same airline for my next journey
20. I prefer airlines which contributes their efforts to·sards carbon free en;·uorunent.
Persuasion (-criteria for chooling a flight)
_: 1. I do not alway:; fly during weekendS-.
.., : I do not alway:; me the same airline my ancestors uses.·used
23. I easily get pers.uadedby cheap tickets: cheap Uckets are at highest priority.
2..:1-. I do not ahvaysget persuaded by more baggage weight allow~d-
25 _ I do not ahvays prefer airlines \Yith best food ser-.-ices; l do not eat much during Journey.
:'6 I do not alv;ayst1y during daytime.
I do not ahvays prefer airlines \;.ith excellent complaint handling services
Qualitative Perception of.Fl<xlbility
28. for me a system is t1exible if I can do changes in my itineraJ"y·.
29. For me a system is. flexible if I cancompletemy\vork.
30. For me a sy·stemis flexible if I can recover from mistakes quickly and easily
31. For me a system is flexible if it offers all features I intend to me.
SatiSfaction
32. O,·erall, I am not satisfied \"\ith how easy it is to use this -....-ebsite.
33. It was not simple tousethliwebsite.
3-t. I cannot effectively complete my,wrk using this \Vebs.ite.
35 I am not able to complete my work quickly using this web rite.
36. I am not able to efficiently complete my ·work us:ing this website.
3--: I do not feel comfortableming this \\"ebsite.
38 Itwunoteasyto learn to use this website.
39. I belieYe- I became unproducti..-e quickly us.ing this websitt!.
--lO. The \Yebsite does not gi...-e error mess.ages that clearly tell me how to fix problems.
--1-1. \\bene,·er, I make a mistake using the website, I cannotreco..-er easily and quickly.
-1-::.. The information {s.uch as online help, on-page messages., and other docwnentation)prtwided ,·vith this. website is not dear.
-H. It is not easy to findtheinformationineed.
-U. The infonnationproYided by the ... vebsite is not eaiy to tn1derstand.
200
- - 0 + ++
- - 0 + ++
- - 0 + ++
- I - 0 + ++
4~ TIH·mfomutionts not effectiYe in helping me complete th~ task~ and scm.mo5
46. The organization ·~finformation on the \•.·ebsitt> pages 1s nN deal
4- Tht> mterf3ce of this. \Yebs:ite ts not pleasant
-1-S I do not lik,;o using the .interface ofthis ·.yebsite
.+9 lhi:. ·xebsite does notha\·e all the function:. and capabilities I expecttt to ha',
50 O:erall. I amnots.atir..tied '"ith this website
Cltf..apFarevs. Your Preferenc-es
51 Ifyou are offered a cheap flight fare, up to >Yhat extent you may compromi:e on the follo,•.ing ~
a I may fly from neighbor to\m rather n~,ing fromm:· O\\TI tO\\TI
b I may compromise on seat comforts. and leg space area
c I may choose a flight '-'ithout jet • . ..-ays
d I rna:· compromise on naffattitude
e I may tly during weekdays
f. I may compromiseonancestryrel.ationships.
g. I may compromise on baggage "\';eight allowed
h I may compromise on food and sen ices
I may fly during night time
I may compromise on complaint handling services.
k I may compromise on airline· s repute
-
:g. Thank you \"ery· much for completing this questionnaire~
201
- 0 + ++
APPENDIX C
USERS' FLEXIBILITY TOWARDS FLEXIBLE ONLINE AIRLINE
RESERVATION SYSTEMS
rs.ers' FlexibilitY tmYards Flexible Online Airline Reservation SYstems (FOARS)
Obj~cthoe - The objective of this questionnaire is to find out the degree of flexibility of airline tra•·elers. The feedback received may help in designing more flexible reservation systems
L Age
Gender
3. Educa tiona! Backgrormd
~ Occupation
5. Ethnicity
General Information (Pluse plu:e a tick() at the appropriate bo:x)
83 Belo•.y .:0 yean EB ::1 to 30 yean
41 to 50 years :51 to60 yean
OJ ::\!ale 0 Female
83 High School Q Diploma
Post-graduate
83 Student m Professional
Ri!tired
8j EU \Vestem
rn AYan Far East
3 1 to-W year:>
Abo.-..-e 60 years
[TI Graduate
[I] Self Employed
[3l African
IT] Others
Rate :rour prioriti~s for th• following sen·i.ce quality IDghest High :Seutral Low attributn of an airline in tenus of their importance.
1 ' 3 • -6 Flymg date confirmation
Fl~,mg canier confirmation 8. Flvlng time confinnation 9 ~wnberofs.top~m:er
10 ~Umber of connected fligha 11 T1cket cla~s (economy business) 12 Seat specifications 13 Discounted airfares H Origin Destination allports 15 Inm.1ediate confirmation ofitinerarv on purchase ofticket
Please indicate which of the foiJo1dng ~tatements. destribes you in the best ponible ,,·ay':"
16. 1 rarely prefer my O>\n country ·s airlme.
1-:-_ !rarely prefer airlines thatpro,idejet \vays, so that I don-t ha\·eto climb the gang.-vay and \\.·ait for busses.
18. I rarely prefer to fly during occasioru
19 Peaktilne tra,·el does not fascinate ahvays
2.0 Off time ttaYel does not fascinate always
21. I \\·auld pn•fer to use systems that conuibute their efforts t0"\\<4tds carbon free envuorunent.
202
1
Least
5
• 5
Staff attitude n.: .. ·ermoti'>'ates.me to m.e the s.am~ airbne
_.:: Irart'lY prefer auhnes. '.'.ith best food serYice5: I do not -eat much Juring joumey
~ 1 Irardy shop. eat or me intemet atthe a.Upon
Cheap tickets are at llighest pnmity in my itmt'rary plarmmg
~ 5 I pr<"f~r onlme ;,ystems because I can tlnd a better package
I rarely get permaded by a hlgher baggage 'c\"E'lght being allo·.;:ed
:8. Ir.uely prefer airbnes with excellent complamthandlingserYices
~ 9 I pre fer manual check-in onr onhne check-m.
3C• I am not a frequent flyer: I t1y only once in a year or so.
3 L I normally t1y to spend holidays \'.ith family friends.
~ ~ I preferto t1y during my children holidays .
.:: J. I normally fly "'i.th my family or friends
34 I notmally flv to expandmy busmess or to meet my professional obJeCti'- es
3 5. Ho\': do you rate your overall personality in terms. of flexibility"
3 6. I am a regular news reader; I always keep mys.elfupdated on tow1sm.
3 ~ I am excellent in managing trip$; I can plan my trip on a Yery illort notlce
3 8 I am excellent in computer s.kil.k I can book my itinerar:_;.· in a ,·er;.· short tnne.
3 9. Online systems are not Yer:_;.· complicated for me to use
40 Online systems proYide support to help recoYer from mistakes
Ptl'teived Usefulness
This website has all the functions. and capabilities I expect 1t to ha\'e
I .. .,.-ill be mcore productiveus.ing this website to book my itinerary
3. I caneffectP:elymakerese:rYations.using this website
4 I am able to book my itinerary· more quickly using thi:; website
It1s easytomakereser.·atiomming this website
6 I am able to efficiently complete my \\·ork using this ·website
I tis easy to learn to me this website
S I feel comfortable u.ting this website
9 It is simple tome this website.
10. The"l.•;ebrite is flexible to interact \\-ith.
11. The mformation u effecttn m helping me complete the tas.b and s.cenanos
12 I find this '>Yebsite easy to me_
13. O.·erall. I am satisfied '"lth the functionality it offers
1-+. Ch·erall. I am satisfied \\ith thi!- website .
. :Q; Thank you nry much for completing tills questionnaire ~
203
1 2 3 4 5
APPENDIX D
TRANSFORMATION OF EVALUATORS USERS' FLEXIBILITY SCORING
Evaluaton
U•tr I Us~r:;
User 3 U•er.: u ... ,~ U;er 6 u,er . U•er S u ... r9 Uier 10
Date Carner Confu:m.anon Confumation
User II .J
u.n 1::: u..,rB User l.J User 15 U"'r 16 Us.....or 1-Uler 1S User 19 User 20 Us."!' 21
U"'r 2! u~er 13 User 1-J User 2)
U•er:'fi u~er r: User 28 User:9 U1er 30 u~er 31 Uler 3::: User 3~
User 3-1 Usa35 Us~r 36 U;er i7
U•er 38 Usn 19 User .. IC' User ~I Uitr 4: u~er 43 U1er.J-l U•tr ~5 User.J6 User-1: User.JS Uier .J9 Um50 User 51 Um~:
Unr 53 UsuS..\ User:'~
User 56 User 5-Usn 5£ Usn 59 U;er 50
Unr51 Usn 62
Traurlonnation of Evaluators [wrs' Fle-xibility Sco1iug
Time Confum.ati~n
:\o_of Stop-o·.·er
' I
' 3
' ' ' '
;.;-o_ of Ccnne~ted -·
' ' ' '
' j
Tickft Cl:n:i
204
Se~t
Specrfu:ation
' '
lmme.:hate Confum11Hm ofltinerarv
o~mnatll>n
Source A» ort
FleXlblhtv Sc<.'re
II 3 s
" II
" ll
6 II
" 16 t: 1-
' IS 10
l8
J: 6 19
' IS ,-1:'
"
" " s
II
" 10 9 J: 10 s 9
9
' s
9
APPENDIX E
CLASSIFICATION OF TRAVELERS' ON THE BASIS OF THEIR FLEXIBLE
TRA YEUNG BEHAVIOR
Classification of Tra\ elen' on the basis of their Flexible Tran•ling Beh:i\"ior
Dlt~ Camer c~nfiunaum Confumaum
U•er 1 U>er: U;tr3 Um-l Um ~ Um6 u .. ,-U•" ~ u .... 9
U<et 10 U<er 11
Us~r 1: u .. , 13 u .. rt-l u .. , !S U!tt 10 Usu 1 Uttr l S u ... 19 Us., :o Us~r :1 Uur :: U;er :J u.n:l
U..-r 26 U:~er :U:;.,l :s u ... :9 U.er30 Um31 Um3: Uoer 33
u,~, 3J u,., 35
u .. , 36 u ... ;u • ..-;s u •• , )9
us~r ~o
U;~r .11
U;~r -12 u ... -13 Um-l-l u ... 4~ u •• , 46 u ... . u ... ~s Uur..l9
~-Um52 u ... ~l u ... ~-~ lloer'' u .... 56 um:U>tr:s User 59 Um0·) Um61 Umo:
:---
Clu~ification:
1 - Lo«. Fl~xible ] - ?\.1ediurr. Fl~xible
'- Htgh Fkx1ble
T~
Confimu.nm
~<> of Xo cf TICk~! Stop. Cmm~ctM Cbu onr fb
-r--- ~ ' -~- ~ --~-~-
--~~ r--+-
205
Sect Specificauon
' '
' ·--~ ---
Dlsccl.llll~d Imm~di2t<
Al:fu~s Confumatial ofhmerar.
'
'
'
-;- --f---~
'-r-j
~uination u..,n· Source Flnlt>llm·
on &No
' r--" ' " " " " " " " " ,-9
" 10
1S
" 6
" " ,-" "
" " 1 1
---
" 13
f::~i 10 9
" 00
• 00 9
:-r--: ---
ClaJ.Iifv
APPENDIX F
PERMISSION TO COLLECT DATA FROM IRA VEL AGENTS FOR RESEARCH
··~S·:'d' '~-~ •1'1', ~-· -~~ :·_;q.: \'ch•~ IL '
<>' ( ·- of! t.:'
PURPOSES
· · t 't-<'-'Pf-t-"11 ·~-p hJ"'-IrHr • H•· ,._ •:~.
-n·-··· ··•··:. Anj >>:n~v _r _. -.u··d··:,-':' ·•·!~-: •, l"l '"I'" /•':)f>
opH~ c' olri\·i '.;H~II,...rir-:_1
-!•.<_ ""·1'-'''1·. ,.,. <_l ,_·!~;" h11
.. ," ··t _..,t d:t n / .. ,, . :·,~rl -~t;dlt:•St•::>;! ~·.til :_.,,_ . !! :0 •:.J;Lil..l·~:.\...::.; . :y·,j.rjpn'L'"I ,)II •,>· •1
1
'N':'•I,;I(:: lik0
··:...<t tH.;:;;-:ot•.o :~.•
"'' --·"':~L
;n .. ,. ' ·'
:·-··
'
··'·· ·.01 •1-·:·1 c_,·,,-?f,-_i'<II'·<J:.c- . .;.:•l!'r•lll:-lr·.-·· ~ ~-1 1>_«~.~8
'lfr),O:Y "'~. ,_-)'-. '<•"-•,Jjf, ~--dt•'ll·"":J:-'J>",-
- •f. tn- ' " ,. ,, .,,.,
'FT ,..li
206
APPENDIX G
PERMISSION TO COLLECT DATA FROM TRAVEL AGENTS AT MA ITA
FAIR20II
',1 . J\ • ,01-r -/•. I 1
Rt:f. P<t:·rrnis!>)OTc io CC<IIEt..::t lJd!l::! from Tnn,•(d Ag'?n1s iit \'!A i T'A r.c~,r 21fC~ ~or P.C'!l!.!-Jrc.:h Purp•!)l!l.fHi
P '\11.~, .. ~;-::J ts o·e c: c !•'•;.· J' _II lfl?,_!·.'::;, Ull '' ~! >J~--
·;;.; .; i ar• s"'"~·-.1:9 ··-r :-·~ w r>·--.-.1 .. ~! '""d': .,- Jy . .Ju:l' d J•~· 'c 1'> .• :
I ,, n ·!
:1•011 i<~:l ·J'lL4 :·ol!;::, ·.:c·n '-'>' ·f .-,_~.,.-; lh•? c.t __ ,j·~ 1-: :.::
•.I"' 1,; ~ ·o:on•: ·r·· ···;rt.·-· ~o con:a-:;1
l t:•u-:: ·:.J:r . .r ~ ·· R·~·- :; - ;-1~---r
D•.·j:,_>\ tl<'!•\' ··:(!'''l~~:<,F 3 .. •) ''o
! J· I.'• ·r,.;l 'r r.~ :.z; ! 'l:... J!·:·:._·,Ni\ ,_ . . :·~ ::..
" . -·· "'
t.l-· I.:._ c :C:C I t•.,. ~->?;.~t · ,-;~rr-::· • ;,
.. • ·•, · ' .·~-t· n: >! I ·· ·
, I
207
!F:Dt'-l:\S ·.·.····,·v~
-;,~ -~,,-<1 :~F!t->Jr nr. Hl'.'t'~·; ,J--d ::ct-?']l(jl
•, · ;o~'l:·.,.., '•'• /•.r r! ~f)
'I I ·: '1) •;::; '.'>'""!';
··r·::i•, ;'JI >::<S+.::l :.r ~,. >l 'I •J !•t :It,..
<·' ;:-,;:; SE' jo:
11.;',;1'-.!' k \;~
APPENDIXH
PERMISSION TO COLLECT DATA FROM TRAVEL AGENTS AT MATTA
FAIR 2011
2.: : 3 : Tuk >•qu•nm
" R"lu"tl" Um' Worldooo
' 26 Do"" R<qu._ U•m co: : R"Ong u .. Um< .'m
' c .. ~'"" '~ .. " •"'•bSi" • '''lu"
'· 61 Dup1•:•
0: ~"' 1mponm<'
63 P>oo:1mport.n< l<:m"' Top C '""'
6.6 Lmgtlu
ll I
' ~ 9.1 u .. 9 u,. Uruqu• md Dmrip:i·. • HnOng:
94 """'' """ ""h 11. u,. B~ok T•x>on ~>m, H:[<h • I
II~ Fo=tCommonlt~msCoruistentlv
II 3 £ruur~ \-~.rual Coruistency
IIA UuBoldT~xtSparingly
II 5 u~~ Anennon-Alnactmg F~atur~' wh..n ,>,ppropnate
II ,15 Un F:urnhu Fonu
II.~ Un at Le:an 1 ~-Pomt Font
ll_S Color-Coding and liHtructiom
119 bnpha~tztlmpN1anco
Cha ter IS: \Yriti \\"eb Content 1 5 1 :\lake ActiCin Sequences Clear
152 U~Famlhu\Yonh
15 3 Ddlne Acionynu andAbbre,"iatiom
I 5 .J Use ).!P;ed Ca~ "ith PrOie
15.5 limit the );"urnber ofWords :and Smtence•
15 6 Use Acti,·e Voice
Cha ter U: Conkrat 0 a1Uution 16 1 Clrg:ln1Zt lnfcunanon C"!nrtv
16.2 Ensw• th:;n );~ceuar.· lnformanonio~Di•played
163 fuou.pReb.tedElements
16.4 ~linimize the ::-ownber ofChcks or PagH
16.': De.sign Qu.;;tntirnon Contont f<:>t Qtu~k Undeuunding
16 6 Di!pbv Only :'\ecus:ot" lnfonnaMn
16 "' F<:>nn~tlnformation for:\lultople Auillencu
16 S Uu Color for Gr"upm~
:'\ote: Assign each Qtlideline a problem Siiv<Htty rattng as follows
NrA "'Not Applicable 0 "' I dent "YH thn tlus i• a u;:otbthty pr"bl~m at •ll
1 "'Ccunencproblentonlv n<:ednctbe r~~edunle~• e.nutinu" "'allableonproJ<~t 2 "' :\I met usablln: ptobl .. m fixmg tlus >hculd be go.ven '""' pn~ntv ] = :\bj"tuubthty ptobl~m- llllpO)tUilll" !1.>:., <O J.h,,uld be g>\'en lugh pn~n" 4 "Unbifuy caum"phe uupeuw·eto fu:tl-u• bd<:>.-e product can be rele:. .. d
208